diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 1de6151f8722647e01f231f6df659f5e0d542a4e..38f068793618d995c2c8fadd750ff597d1562b4c 100644 --- a/analysis/current/analysis-report-hybrid.json +++ b/analysis/current/analysis-report-hybrid.json @@ -1,2101 +1,3117 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260721T180022Z", - "generated_at": "2026-07-21T18:20:54Z", + "snapshot_id": "20260722T000038Z", + "generated_at": "2026-07-22T00:15:44Z", "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.", + "cluster_id": "cluster-47122-13", + "summary": "Cluster of 13 related pull requests linked by soft_similarity.", "status": "open", - "confidence": 0.8, - "canonical_issue_number": 43979, - "canonical_pr_number": 44007, - "issue_numbers": [ - 43979 - ], + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47122, + "issue_numbers": [], "pr_numbers": [ - 43996, - 44007, - 44013, - 44044, - 44066, - 44072, - 44085, - 44129, - 44154, - 44722 + 47122, + 47123, + 47124, + 47224, + 47266, + 47280, + 47281, + 47282, + 47310, + 47380, + 47443, + 47444, + 47445 ], "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, + "left_pr_number": 47122, + "right_pr_number": 47123, + "code_similarity": 0.092, + "size_similarity": 0.462, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.408, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 43996, - "right_pr_number": 44072, - "code_similarity": 0.14, - "size_similarity": 0.303, + "left_pr_number": 47122, + "right_pr_number": 47124, + "code_similarity": 0.2, + "size_similarity": 1.0, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.528, + "patch_similarity": 0.0, "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": 47122, + "right_pr_number": 47224, + "code_similarity": 0.474, + "size_similarity": 0.302, + "file_overlap": 0.4, + "area_overlap": 0.818, + "patch_similarity": 0.051, + "shared_filenames": [ + "src/transformers/models/edgetam/modeling_edgetam.py", + "src/transformers/models/edgetam/modular_edgetam.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/edgetam/modeling_edgetam.py", + "left_ranges": [ + [ + 464, + 472 + ] + ], + "right_ranges": [ + [ + 464, + 474 + ] + ] + }, + { + "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "left_ranges": [ + [ + 210, + 218 + ] + ], + "right_ranges": [ + [ + 210, + 220 + ] + ] + } + ] }, { - "left_pr_number": 43996, - "right_pr_number": 44129, - "code_similarity": 0.163, - "size_similarity": 0.643, + "left_pr_number": 47122, + "right_pr_number": 47266, + "code_similarity": 0.169, + "size_similarity": 0.846, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.229, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 43996, - "right_pr_number": 44154, - "code_similarity": 0.153, - "size_similarity": 0.535, + "left_pr_number": 47122, + "right_pr_number": 47280, + "code_similarity": 0.177, + "size_similarity": 0.885, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.31, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 43996, - "right_pr_number": 44722, - "code_similarity": 0.225, - "size_similarity": 0.848, + "left_pr_number": 47122, + "right_pr_number": 47281, + "code_similarity": 0.05, + "size_similarity": 0.228, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.368, + "patch_similarity": 0.03, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44013, - "code_similarity": 0.19, - "size_similarity": 0.553, + "left_pr_number": 47122, + "right_pr_number": 47282, + "code_similarity": 0.118, + "size_similarity": 0.591, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.531, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44044, - "code_similarity": 0.186, - "size_similarity": 0.667, + "left_pr_number": 47122, + "right_pr_number": 47310, + "code_similarity": 0.111, + "size_similarity": 0.5, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.354, + "patch_similarity": 0.074, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44066, - "code_similarity": 0.188, - "size_similarity": 0.704, + "left_pr_number": 47122, + "right_pr_number": 47380, + "code_similarity": 0.169, + "size_similarity": 0.846, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.315, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44072, - "code_similarity": 0.212, - "size_similarity": 0.526, + "left_pr_number": 47122, + "right_pr_number": 47443, + "code_similarity": 0.011, + "size_similarity": 0.051, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.708, + "patch_similarity": 0.006, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44085, - "code_similarity": 0.195, - "size_similarity": 0.735, + "left_pr_number": 47122, + "right_pr_number": 47444, + "code_similarity": 0.02, + "size_similarity": 0.08, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.318, + "patch_similarity": 0.023, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44129, - "code_similarity": 0.103, - "size_similarity": 0.37, + "left_pr_number": 47122, + "right_pr_number": 47445, + "code_similarity": 0.02, + "size_similarity": 0.08, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.191, + "patch_similarity": 0.025, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44154, - "code_similarity": 0.238, - "size_similarity": 0.93, + "left_pr_number": 47123, + "right_pr_number": 47124, + "code_similarity": 0.092, + "size_similarity": 0.462, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.344, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44007, - "right_pr_number": 44722, - "code_similarity": 0.178, - "size_similarity": 0.679, + "left_pr_number": 47123, + "right_pr_number": 47224, + "code_similarity": 0.028, + "size_similarity": 0.14, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.28, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44044, - "code_similarity": 0.126, - "size_similarity": 0.368, + "left_pr_number": 47123, + "right_pr_number": 47266, + "code_similarity": 0.159, + "size_similarity": 0.545, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.351, + "patch_similarity": 0.333, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44066, - "code_similarity": 0.127, - "size_similarity": 0.389, + "left_pr_number": 47123, + "right_pr_number": 47280, + "code_similarity": 0.547, + "size_similarity": 0.522, + "file_overlap": 1.0, + "area_overlap": 0.368, + "patch_similarity": 0.093, + "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, + 21 + ], + [ + 481, + 499 + ], + [ + 520, + 530 + ] + ] + } + ] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47281, + "code_similarity": 0.023, + "size_similarity": 0.105, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.325, + "patch_similarity": 0.011, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44072, - "code_similarity": 0.29, - "size_similarity": 0.952, + "left_pr_number": 47123, + "right_pr_number": 47282, + "code_similarity": 0.055, + "size_similarity": 0.273, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.667, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44085, - "code_similarity": 0.131, - "size_similarity": 0.406, + "left_pr_number": 47123, + "right_pr_number": 47310, + "code_similarity": 0.185, + "size_similarity": 0.923, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.329, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44129, - "code_similarity": 0.07, - "size_similarity": 0.205, + "left_pr_number": 47123, + "right_pr_number": 47380, + "code_similarity": 0.109, + "size_similarity": 0.545, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.192, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44154, - "code_similarity": 0.177, - "size_similarity": 0.594, + "left_pr_number": 47123, + "right_pr_number": 47443, + "code_similarity": 0.006, + "size_similarity": 0.023, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.389, + "patch_similarity": 0.01, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44013, - "right_pr_number": 44722, - "code_similarity": 0.118, - "size_similarity": 0.375, + "left_pr_number": 47123, + "right_pr_number": 47444, + "code_similarity": 0.009, + "size_similarity": 0.037, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.287, + "patch_similarity": 0.013, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44066, - "code_similarity": 0.25, - "size_similarity": 0.947, + "left_pr_number": 47123, + "right_pr_number": 47445, + "code_similarity": 0.01, + "size_similarity": 0.037, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.404, + "patch_similarity": 0.014, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44072, - "code_similarity": 0.136, - "size_similarity": 0.351, + "left_pr_number": 47124, + "right_pr_number": 47224, + "code_similarity": 0.06, + "size_similarity": 0.302, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.442, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44085, - "code_similarity": 0.24, - "size_similarity": 0.906, + "left_pr_number": 47124, + "right_pr_number": 47266, + "code_similarity": 0.169, + "size_similarity": 0.846, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.394, + "patch_similarity": 0.0, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44129, - "code_similarity": 0.147, - "size_similarity": 0.555, + "left_pr_number": 47124, + "right_pr_number": 47280, + "code_similarity": 0.179, + "size_similarity": 0.885, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.243, + "patch_similarity": 0.013, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44154, - "code_similarity": 0.17, - "size_similarity": 0.62, + "left_pr_number": 47124, + "right_pr_number": 47281, + "code_similarity": 0.048, + "size_similarity": 0.228, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.306, + "patch_similarity": 0.016, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44044, - "right_pr_number": 44722, - "code_similarity": 0.257, - "size_similarity": 0.982, + "left_pr_number": 47124, + "right_pr_number": 47282, + "code_similarity": 0.128, + "size_similarity": 0.591, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.402, + "patch_similarity": 0.067, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 44066, - "right_pr_number": 44072, - "code_similarity": 0.133, - "size_similarity": 0.37, + "left_pr_number": 47124, + "right_pr_number": 47310, + "code_similarity": 0.1, + "size_similarity": 0.5, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.393, + "patch_similarity": 0.0, "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 + "left_pr_number": 47124, + "right_pr_number": 47380, + "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": 47124, + "right_pr_number": 47443, + "code_similarity": 0.014, + "size_similarity": 0.051, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.025, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47444, + "code_similarity": 0.016, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47445, + "code_similarity": 0.016, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47266, + "code_similarity": 0.056, + "size_similarity": 0.256, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47280, + "code_similarity": 0.058, + "size_similarity": 0.267, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47281, + "code_similarity": 0.159, + "size_similarity": 0.754, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.052, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47282, + "code_similarity": 0.105, + "size_similarity": 0.512, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.018, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47310, + "code_similarity": 0.037, + "size_similarity": 0.151, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.047, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47380, + "code_similarity": 0.056, + "size_similarity": 0.256, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.032, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47443, + "code_similarity": 0.039, + "size_similarity": 0.168, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47444, + "code_similarity": 0.058, + "size_similarity": 0.266, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47445, + "code_similarity": 0.058, + "size_similarity": 0.266, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47280, + "code_similarity": 0.229, + "size_similarity": 0.957, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.25, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47281, + "code_similarity": 0.045, + "size_similarity": 0.193, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.04, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47282, + "code_similarity": 0.104, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.028, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47310, + "code_similarity": 0.118, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47380, + "code_similarity": 0.226, + "size_similarity": 1.0, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.174, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47443, + "code_similarity": 0.177, + "size_similarity": 0.043, + "file_overlap": 0.1, + "area_overlap": 0.387, + "patch_similarity": 0.02, + "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, + 20 + ], + [ + 245, + 264 + ], + [ + 277, + 283 + ], + [ + 298, + 304 + ], + [ + 339, + 345 + ] + ], + "right_ranges": [ + [ + 13, + 19 + ], + [ + 245, + 360 + ] + ] + } + ] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47444, + "code_similarity": 0.019, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47445, + "code_similarity": 0.021, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.052, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47281, + "code_similarity": 0.046, + "size_similarity": 0.202, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47282, + "code_similarity": 0.107, + "size_similarity": 0.523, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.02, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47310, + "code_similarity": 0.116, + "size_similarity": 0.565, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.017, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47380, + "code_similarity": 0.205, + "size_similarity": 0.957, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.094, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47443, + "code_similarity": 0.014, + "size_similarity": 0.045, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.036, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47444, + "code_similarity": 0.018, + "size_similarity": 0.071, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47445, + "code_similarity": 0.02, + "size_similarity": 0.071, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.037, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47282, + "code_similarity": 0.082, + "size_similarity": 0.386, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.034, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47310, + "code_similarity": 0.026, + "size_similarity": 0.114, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.019, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47380, + "code_similarity": 0.046, + "size_similarity": 0.193, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.049, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47443, + "code_similarity": 0.052, + "size_similarity": 0.223, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.051, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47444, + "code_similarity": 0.082, + "size_similarity": 0.353, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.076, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47445, + "code_similarity": 0.084, + "size_similarity": 0.353, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.088, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47310, + "code_similarity": 0.063, + "size_similarity": 0.295, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47380, + "code_similarity": 0.476, + "size_similarity": 0.5, + "file_overlap": 1.0, + "area_overlap": 0.194, + "patch_similarity": 0.056, + "shared_filenames": [ + "tests/models/mamba2/test_modeling_mamba2.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/mamba2/test_modeling_mamba2.py", + "left_ranges": [ + [ + 420, + 438 + ], + [ + 450, + 474 + ], + [ + 482, + 524 + ] + ], + "right_ranges": [ + [ + 18, + 24 + ], + [ + 406, + 419 + ], + [ + 426, + 432 + ], + [ + 457, + 463 + ], + [ + 487, + 493 + ] + ] + } + ] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47443, + "code_similarity": 0.027, + "size_similarity": 0.086, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.068, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47444, + "code_similarity": 0.032, + "size_similarity": 0.136, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.03, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47445, + "code_similarity": 0.033, + "size_similarity": 0.136, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.04, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47380, + "code_similarity": 0.123, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.032, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47443, + "code_similarity": 0.008, + "size_similarity": 0.025, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.023, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47444, + "code_similarity": 0.015, + "size_similarity": 0.04, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.044, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47445, + "code_similarity": 0.015, + "size_similarity": 0.04, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.049, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47443, + "code_similarity": 0.013, + "size_similarity": 0.043, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.03, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47444, + "code_similarity": 0.02, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.045, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47445, + "code_similarity": 0.023, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.063, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47443, + "right_pr_number": 47444, + "code_similarity": 0.754, + "size_similarity": 0.632, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.217, + "shared_filenames": [ + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "left_ranges": [ + [ + 4, + 9 ], [ - 397, - 405 + 310, + 315 + ] + ], + "right_ranges": [ + [ + 4, + 9 ], [ - 408, - 419 + 310, + 315 + ] + ] + }, + { + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "left_ranges": [ + [ + 299, + 304 ], [ - 425, - 434 + 338, + 349 ], [ - 459, - 465 + 392, + 403 + ] + ], + "right_ranges": [ + [ + 299, + 304 ], [ - 470, - 489 + 338, + 349 ], [ - 517, - 539 + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ + [ + 182, + 187 ], [ - 553, - 559 + 221, + 232 ], [ - 565, - 574 + 272, + 283 + ] + ], + "right_ranges": [ + [ + 182, + 187 ], [ - 579, - 597 + 221, + 232 ], [ - 600, - 611 + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ + [ + 306, + 311 ], [ - 633, - 639 + 358, + 369 + ] + ], + "right_ranges": [ + [ + 306, + 311 ], [ - 645, - 652 + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ], + "right_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ] + }, + { + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "left_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ], + "right_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ], + "right_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 ], [ - 657, - 673 + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ] + }, + { + "filename": "src/transformers/models/youtu/modeling_youtu.py", + "left_ranges": [ + [ + 222, + 227 ], [ - 716, - 728 + 261, + 272 ], [ - 737, - 743 + 312, + 323 + ] + ], + "right_ranges": [ + [ + 222, + 227 ], [ - 748, - 771 + 261, + 272 ], [ - 789, - 794 + 312, + 323 ] ] } ] }, { - "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, + "left_pr_number": 47443, + "right_pr_number": 47445, + "code_similarity": 0.755, + "size_similarity": 0.632, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.219, "shared_filenames": [ - "src/transformers/models/codegen/modeling_codegen.py", - "src/transformers/models/gptj/modeling_gptj.py" + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" ], "shared_file_areas": [ { - "filename": "src/transformers/models/codegen/modeling_codegen.py", + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", "left_ranges": [ [ - 245, - 266 + 4, + 9 + ], + [ + 310, + 315 ] ], "right_ranges": [ [ - 228, - 234 + 4, + 9 + ], + [ + 310, + 315 ] ] }, { - "filename": "src/transformers/models/gptj/modeling_gptj.py", + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", "left_ranges": [ [ - 33, - 41 + 299, + 304 ], [ - 174, - 181 + 338, + 349 ], [ - 250, - 257 - ], + 392, + 403 + ] + ], + "right_ranges": [ [ - 398, - 419 + 299, + 304 ], [ - 425, - 434 + 338, + 349 ], [ - 459, - 466 - ], + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ [ - 471, - 488 + 182, + 187 ], [ - 518, - 540 + 221, + 232 ], [ - 554, - 560 - ], + 272, + 283 + ] + ], + "right_ranges": [ [ - 566, - 575 + 182, + 187 ], [ - 580, - 598 + 221, + 232 ], [ - 601, - 606 + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ + [ + 306, + 311 ], [ - 634, - 640 + 358, + 369 + ] + ], + "right_ranges": [ + [ + 306, + 311 ], [ - 646, - 653 + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ + [ + 282, + 287 ], [ - 658, - 674 + 321, + 332 ], [ - 717, - 722 + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 ], [ - 738, - 744 + 321, + 332 ], [ - 749, - 772 + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 ], [ - 790, - 795 + 283, + 288 ] ], "right_ranges": [ [ - 33, - 42 + 196, + 201 ], [ - 175, - 182 - ], + 283, + 288 + ] + ] + }, + { + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "left_ranges": [ [ - 251, - 258 + 17, + 22 ], [ - 399, - 420 + 319, + 324 + ] + ], + "right_ranges": [ + [ + 17, + 22 ], [ - 426, - 435 + 319, + 324 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 109, + 114 ], [ - 460, - 467 + 291, + 300 ], [ - 472, - 489 + 361, + 370 ], [ - 519, - 540 + 396, + 409 ], [ - 554, - 560 - ], + 435, + 444 + ] + ], + "right_ranges": [ [ - 566, - 575 + 109, + 114 ], [ - 580, - 598 + 291, + 300 ], [ - 601, - 612 + 361, + 370 ], [ - 634, - 640 + 396, + 409 ], [ - 646, - 653 - ], + 435, + 444 + ] + ] + }, + { + "filename": "src/transformers/models/youtu/modeling_youtu.py", + "left_ranges": [ [ - 658, - 663 + 222, + 227 ], [ - 666, - 674 + 261, + 272 ], [ - 717, - 722 - ], + 312, + 323 + ] + ], + "right_ranges": [ [ - 738, - 744 + 222, + 227 ], [ - 749, - 772 + 261, + 272 ], [ - 790, - 795 + 312, + 323 ] ] } ] }, { - "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, + "left_pr_number": 47444, + "right_pr_number": 47445, + "code_similarity": 0.916, + "size_similarity": 1.0, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.802, "shared_filenames": [ - "src/transformers/models/gptj/modeling_gptj.py" + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" ], "shared_file_areas": [ { - "filename": "src/transformers/models/gptj/modeling_gptj.py", + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.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 + 4, + 9 ], [ - 645, - 652 - ], + 310, + 315 + ] + ], + "right_ranges": [ [ - 657, - 673 + 4, + 9 ], [ - 716, - 728 - ], + 310, + 315 + ] + ] + }, + { + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "left_ranges": [ [ - 737, - 743 + 299, + 304 ], [ - 748, - 771 + 338, + 349 ], [ - 789, - 794 + 392, + 403 ] ], "right_ranges": [ [ - 33, - 42 + 299, + 304 ], [ - 175, - 182 + 338, + 349 ], [ - 251, - 258 - ], + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ [ - 399, - 420 + 182, + 187 ], [ - 426, - 435 + 221, + 232 ], [ - 460, - 467 - ], + 272, + 283 + ] + ], + "right_ranges": [ [ - 472, - 489 + 182, + 187 ], [ - 519, - 540 + 221, + 232 ], [ - 554, - 560 - ], + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ [ - 566, - 575 + 306, + 311 ], [ - 580, - 598 - ], + 358, + 369 + ] + ], + "right_ranges": [ [ - 601, - 612 + 306, + 311 ], [ - 634, - 640 - ], + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ [ - 646, - 653 + 282, + 287 ], [ - 658, - 663 + 321, + 332 ], [ - 666, - 674 + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 ], [ - 717, - 722 + 321, + 332 ], [ - 738, - 744 + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 ], [ - 749, - 772 + 283, + 288 + ] + ], + "right_ranges": [ + [ + 196, + 201 ], [ - 790, - 795 + 283, + 288 ] ] - } - ] - }, - { - "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-47122-11", - "summary": "Cluster of 11 related pull requests linked by soft_similarity.", - "status": "open", - "confidence": 0.5, - "canonical_issue_number": null, - "canonical_pr_number": 47122, - "issue_numbers": [], - "pr_numbers": [ - 47122, - 47123, - 47224, - 47266, - 47280, - 47281, - 47282, - 47310, - 47380, - 47443, - 47444 - ], - "evidence_types": [ - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 47122, - "right_pr_number": 47123, - "code_similarity": 0.092, - "size_similarity": 0.462, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47224, - "code_similarity": 0.474, - "size_similarity": 0.302, - "file_overlap": 0.4, - "area_overlap": 0.818, - "patch_similarity": 0.051, - "shared_filenames": [ - "src/transformers/models/edgetam/modeling_edgetam.py", - "src/transformers/models/edgetam/modular_edgetam.py" - ], - "shared_file_areas": [ + }, { - "filename": "src/transformers/models/edgetam/modeling_edgetam.py", + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", "left_ranges": [ [ - 464, - 472 + 17, + 22 + ], + [ + 319, + 324 ] ], "right_ranges": [ [ - 464, - 474 + 17, + 22 + ], + [ + 319, + 324 ] ] }, { - "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", "left_ranges": [ [ - 210, - 218 + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 ] ], "right_ranges": [ [ - 210, - 220 + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 ] ] - } - ] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47266, - "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": 47122, - "right_pr_number": 47280, - "code_similarity": 0.177, - "size_similarity": 0.885, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47281, - "code_similarity": 0.05, - "size_similarity": 0.228, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.03, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47282, - "code_similarity": 0.118, - "size_similarity": 0.591, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47310, - "code_similarity": 0.111, - "size_similarity": 0.5, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.074, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47380, - "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": 47122, - "right_pr_number": 47443, - "code_similarity": 0.011, - "size_similarity": 0.051, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.006, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47122, - "right_pr_number": 47444, - "code_similarity": 0.02, - "size_similarity": 0.08, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.023, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47123, - "right_pr_number": 47224, - "code_similarity": 0.028, - "size_similarity": 0.14, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.0, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47123, - "right_pr_number": 47266, - "code_similarity": 0.159, - "size_similarity": 0.545, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.333, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47123, - "right_pr_number": 47280, - "code_similarity": 0.547, - "size_similarity": 0.522, - "file_overlap": 1.0, - "area_overlap": 0.368, - "patch_similarity": 0.093, - "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", + "filename": "src/transformers/models/youtu/modeling_youtu.py", "left_ranges": [ [ - 14, - 20 + 222, + 227 ], [ - 487, - 493 + 261, + 272 + ], + [ + 312, + 323 ] ], "right_ranges": [ [ - 14, - 21 + 222, + 227 ], [ - 481, - 499 + 261, + 272 ], [ - 520, - 530 + 312, + 323 ] ] } ] - }, + } + ] + }, + { + "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": 47123, - "right_pr_number": 47281, - "code_similarity": 0.023, - "size_similarity": 0.105, + "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": 0.011, + "patch_similarity": 0.429, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47123, - "right_pr_number": 47282, - "code_similarity": 0.055, - "size_similarity": 0.273, + "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": 0.0, + "patch_similarity": 0.392, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47123, - "right_pr_number": 47310, - "code_similarity": 0.185, - "size_similarity": 0.923, + "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": 0.0, + "patch_similarity": 0.479, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47123, - "right_pr_number": 47380, - "code_similarity": 0.109, - "size_similarity": 0.545, + "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": 0.0, + "patch_similarity": 0.408, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47123, - "right_pr_number": 47443, - "code_similarity": 0.006, - "size_similarity": 0.023, + "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": 0.01, + "patch_similarity": 0.528, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47123, - "right_pr_number": 47444, - "code_similarity": 0.009, - "size_similarity": 0.037, + "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": 0.013, + "patch_similarity": 0.398, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47266, - "code_similarity": 0.056, - "size_similarity": 0.256, + "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": 0.033, + "patch_similarity": 0.229, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47280, - "code_similarity": 0.058, - "size_similarity": 0.267, + "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": 0.033, + "patch_similarity": 0.31, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47281, - "code_similarity": 0.159, - "size_similarity": 0.754, + "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": 0.052, + "patch_similarity": 0.368, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47282, - "code_similarity": 0.105, - "size_similarity": 0.512, + "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": 0.018, + "patch_similarity": 0.531, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47310, - "code_similarity": 0.037, - "size_similarity": 0.151, + "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": 0.047, + "patch_similarity": 0.354, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47380, - "code_similarity": 0.056, - "size_similarity": 0.256, + "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": 0.032, + "patch_similarity": 0.315, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47443, - "code_similarity": 0.039, - "size_similarity": 0.168, + "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": 0.033, + "patch_similarity": 0.708, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47224, - "right_pr_number": 47444, - "code_similarity": 0.058, - "size_similarity": 0.266, + "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": 0.033, + "patch_similarity": 0.318, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47280, - "code_similarity": 0.229, - "size_similarity": 0.957, + "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": 0.25, + "patch_similarity": 0.191, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47281, - "code_similarity": 0.045, - "size_similarity": 0.193, + "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": 0.04, + "patch_similarity": 0.344, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47282, - "code_similarity": 0.104, - "size_similarity": 0.5, + "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": 0.028, + "patch_similarity": 0.28, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47310, - "code_similarity": 0.118, - "size_similarity": 0.591, + "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": 0.0, + "patch_similarity": 0.351, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47380, - "code_similarity": 0.226, - "size_similarity": 1.0, + "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": 0.174, + "patch_similarity": 0.325, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47266, - "right_pr_number": 47443, - "code_similarity": 0.177, - "size_similarity": 0.043, - "file_overlap": 0.1, - "area_overlap": 0.387, - "patch_similarity": 0.02, - "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, - 20 - ], - [ - 245, - 264 - ], - [ - 277, - 283 - ], - [ - 298, - 304 - ], - [ - 339, - 345 - ] - ], - "right_ranges": [ - [ - 13, - 19 - ], - [ - 245, - 360 - ] - ] - } - ] + "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": 47266, - "right_pr_number": 47444, - "code_similarity": 0.019, - "size_similarity": 0.068, + "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": 0.035, + "patch_similarity": 0.329, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47281, - "code_similarity": 0.046, - "size_similarity": 0.202, + "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": 0.038, + "patch_similarity": 0.192, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47282, - "code_similarity": 0.107, - "size_similarity": 0.523, + "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": 0.02, + "patch_similarity": 0.389, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47310, - "code_similarity": 0.116, - "size_similarity": 0.565, + "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": 0.017, + "patch_similarity": 0.287, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47380, - "code_similarity": 0.205, - "size_similarity": 0.957, + "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": 0.094, + "patch_similarity": 0.404, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47443, - "code_similarity": 0.014, - "size_similarity": 0.045, + "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": 0.036, + "patch_similarity": 0.442, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47280, - "right_pr_number": 47444, - "code_similarity": 0.018, - "size_similarity": 0.071, + "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": 0.026, + "patch_similarity": 0.394, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47281, - "right_pr_number": 47282, - "code_similarity": 0.082, - "size_similarity": 0.386, + "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": 0.034, + "patch_similarity": 0.243, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47281, - "right_pr_number": 47310, - "code_similarity": 0.026, - "size_similarity": 0.114, + "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": 0.019, + "patch_similarity": 0.306, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47281, - "right_pr_number": 47380, - "code_similarity": 0.046, - "size_similarity": 0.193, + "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": 0.049, + "patch_similarity": 0.402, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47281, - "right_pr_number": 47443, - "code_similarity": 0.052, - "size_similarity": 0.223, + "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": 0.051, + "patch_similarity": 0.393, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47281, - "right_pr_number": 47444, - "code_similarity": 0.082, - "size_similarity": 0.353, + "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": 0.076, + "patch_similarity": 0.263, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47282, - "right_pr_number": 47310, - "code_similarity": 0.063, - "size_similarity": 0.295, + "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": 0.026, + "patch_similarity": 0.286, "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47282, - "right_pr_number": 47380, - "code_similarity": 0.476, - "size_similarity": 0.5, + "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": 0.194, - "patch_similarity": 0.056, + "area_overlap": 0.808, + "patch_similarity": 0.874, "shared_filenames": [ - "tests/models/mamba2/test_modeling_mamba2.py" + "src/transformers/models/codegen/modeling_codegen.py", + "src/transformers/models/gptj/modeling_gptj.py" ], "shared_file_areas": [ { - "filename": "tests/models/mamba2/test_modeling_mamba2.py", + "filename": "src/transformers/models/codegen/modeling_codegen.py", "left_ranges": [ [ - 420, - 438 + 245, + 266 + ] + ], + "right_ranges": [ + [ + 228, + 234 + ] + ] + }, + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 41 ], [ - 450, - 474 + 174, + 181 ], [ - 482, - 524 + 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": [ [ - 18, - 24 + 33, + 42 ], [ - 406, - 419 + 175, + 182 + ], + [ + 251, + 258 + ], + [ + 399, + 420 ], [ 426, - 432 + 435 ], [ - 457, - 463 + 460, + 467 ], [ - 487, - 493 + 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": 47282, - "right_pr_number": 47443, - "code_similarity": 0.027, - "size_similarity": 0.086, - "file_overlap": 0.0, - "area_overlap": 0.0, - "patch_similarity": 0.068, - "shared_filenames": [], - "shared_file_areas": [] - }, - { - "left_pr_number": 47282, - "right_pr_number": 47444, - "code_similarity": 0.032, - "size_similarity": 0.136, + "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": 0.03, + "patch_similarity": 0.398, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47310, - "right_pr_number": 47380, - "code_similarity": 0.123, - "size_similarity": 0.591, + "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": 0.032, + "patch_similarity": 0.231, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47310, - "right_pr_number": 47443, - "code_similarity": 0.008, - "size_similarity": 0.025, + "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": 0.023, + "patch_similarity": 0.414, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47310, - "right_pr_number": 47444, - "code_similarity": 0.015, - "size_similarity": 0.04, + "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": 0.044, + "patch_similarity": 0.347, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47380, - "right_pr_number": 47443, - "code_similarity": 0.013, - "size_similarity": 0.043, + "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": 0.03, + "patch_similarity": 0.272, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47380, - "right_pr_number": 47444, - "code_similarity": 0.02, - "size_similarity": 0.068, + "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": 0.045, + "patch_similarity": 0.289, "shared_filenames": [], "shared_file_areas": [] }, { - "left_pr_number": 47443, - "right_pr_number": 47444, - "code_similarity": 0.754, - "size_similarity": 0.632, - "file_overlap": 0.818, - "area_overlap": 1.0, - "patch_similarity": 0.217, + "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/deepseek_v3/modeling_deepseek_v3.py", - "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", - "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", - "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", - "src/transformers/models/longcat_flash/modeling_longcat_flash.py", - "src/transformers/models/minicpm3/modeling_minicpm3.py", - "src/transformers/models/mistral4/modeling_mistral4.py", - "src/transformers/models/nemotron_h/modeling_nemotron_h.py", - "src/transformers/models/youtu/modeling_youtu.py" + "src/transformers/models/gptj/modeling_gptj.py" ], "shared_file_areas": [ { - "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "filename": "src/transformers/models/gptj/modeling_gptj.py", "left_ranges": [ [ - 4, - 9 + 33, + 40 ], [ - 310, - 315 - ] - ], - "right_ranges": [ + 173, + 180 + ], [ - 4, - 9 + 249, + 256 ], [ - 310, - 315 - ] - ] - }, - { - "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", - "left_ranges": [ + 397, + 405 + ], [ - 299, - 304 + 408, + 419 ], [ - 338, - 349 + 425, + 434 ], [ - 392, - 403 - ] - ], - "right_ranges": [ + 459, + 465 + ], [ - 299, - 304 + 470, + 489 ], [ - 338, - 349 + 517, + 539 ], [ - 392, - 403 - ] - ] - }, - { - "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", - "left_ranges": [ + 553, + 559 + ], [ - 182, - 187 + 565, + 574 ], [ - 221, - 232 + 579, + 597 ], [ - 272, - 283 - ] - ], - "right_ranges": [ + 600, + 611 + ], [ - 182, - 187 + 633, + 639 ], [ - 221, - 232 + 645, + 652 ], [ - 272, - 283 - ] - ] - }, - { - "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", - "left_ranges": [ + 657, + 673 + ], [ - 306, - 311 + 716, + 728 ], [ - 358, - 369 + 737, + 743 + ], + [ + 748, + 771 + ], + [ + 789, + 794 ] ], "right_ranges": [ [ - 306, - 311 + 33, + 42 ], [ - 358, - 369 - ] - ] - }, - { - "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", - "left_ranges": [ + 175, + 182 + ], [ - 282, - 287 + 251, + 258 ], [ - 321, - 332 + 399, + 420 ], [ - 372, - 383 - ] - ], - "right_ranges": [ + 426, + 435 + ], [ - 282, - 287 + 460, + 467 ], [ - 321, - 332 + 472, + 489 ], [ - 372, - 383 - ] - ] - }, - { - "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", - "left_ranges": [ + 519, + 540 + ], [ - 196, - 201 + 554, + 560 ], [ - 283, - 288 - ] - ], - "right_ranges": [ + 566, + 575 + ], [ - 196, - 201 + 580, + 598 ], [ - 283, - 288 + 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-43451-5", + "summary": "Cluster of 5 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, + 45157, + 46509, + 47163, + 47306 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43451, + "right_pr_number": 45157, + "code_similarity": 0.016, + "size_similarity": 0.036, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.058, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43451, + "right_pr_number": 46509, + "code_similarity": 0.138, + "size_similarity": 0.363, + "file_overlap": 0.097, + "area_overlap": 0.0, + "patch_similarity": 0.243, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ { - "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "filename": "docs/source/en/_toctree.yml", "left_ranges": [ [ - 17, - 22 - ], - [ - 319, - 324 + 1381, + 1388 ] ], "right_ranges": [ [ - 17, - 22 - ], - [ - 319, - 324 + 795, + 802 ] ] }, { - "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "filename": "src/transformers/models/auto/auto_mappings.py", "left_ranges": [ [ - 109, - 114 + 398, + 405 ], [ - 291, - 300 + 841, + 847 ], [ - 361, - 370 + 965, + 971 ], [ - 396, - 409 + 1089, + 1095 ], [ - 435, - 444 + 1225, + 1231 ] ], "right_ranges": [ [ - 109, - 114 - ], - [ - 291, - 300 - ], - [ - 361, - 370 + 412, + 418 ], [ - 396, - 409 + 1077, + 1083 ], [ - 435, - 444 + 1206, + 1212 ] ] }, { - "filename": "src/transformers/models/youtu/modeling_youtu.py", + "filename": "src/transformers/models/auto/modeling_auto.py", "left_ranges": [ [ - 222, - 227 + 335, + 342 ], [ - 261, - 272 + 1107, + 1113 ], [ - 312, - 323 + 1930, + 1936 ] ], "right_ranges": [ [ - 222, - 227 - ], - [ - 261, - 272 - ], - [ - 312, - 323 + 1097, + 1103 ] ] } ] + }, + { + "left_pr_number": 43451, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.004, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.02, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43451, + "right_pr_number": 47306, + "code_similarity": 0.04, + "size_similarity": 0.122, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.107, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 46509, + "code_similarity": 0.032, + "size_similarity": 0.101, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.08, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 47163, + "code_similarity": 0.029, + "size_similarity": 0.117, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 47306, + "code_similarity": 0.078, + "size_similarity": 0.298, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.125, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47163, + "code_similarity": 0.006, + "size_similarity": 0.012, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.024, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47306, + "code_similarity": 0.086, + "size_similarity": 0.337, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.126, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47163, + "right_pr_number": 47306, + "code_similarity": 0.011, + "size_similarity": 0.035, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] } ] }, @@ -2183,14 +3199,109 @@ "filename": "tests/models/deimv2/test_modeling_deimv2.py", "left_ranges": [ [ - 1, - 15 + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 1734 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46894-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 46894, + "canonical_pr_number": 47090, + "issue_numbers": [ + 46894, + 46897 + ], + "pr_numbers": [ + 46958, + 47090 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46958, + "right_pr_number": 47090, + "code_similarity": 0.397, + "size_similarity": 0.257, + "file_overlap": 0.273, + "area_overlap": 0.598, + "patch_similarity": 0.362, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "tests/models/cohere_asr/test_modeling_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ], + "right_ranges": [ + [ + 637, + 649 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ], + "right_ranges": [ + [ + 500, + 512 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 15, + 21 + ], + [ + 180, + 221 ] ], "right_ranges": [ [ - 1, - 1734 + 175, + 224 ] ] } @@ -2199,62 +3310,35 @@ ] }, { - "cluster_id": "cluster-44843-4", - "summary": "Cluster of 2 issues and 2 PRs centered on issue #45081.", + "cluster_id": "cluster-47301-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #47333.", "status": "open", - "confidence": 0.8, - "canonical_issue_number": 45081, - "canonical_pr_number": 45317, + "confidence": 0.6, + "canonical_issue_number": 47333, + "canonical_pr_number": 47336, "issue_numbers": [ - 44843, - 45081 + 47332, + 47333 ], "pr_numbers": [ - 45086, - 45317 + 47301, + 47336 ], "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 - ] - ] - } - ] + "left_pr_number": 47301, + "right_pr_number": 47336, + "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": [] } ] }, @@ -2358,93 +3442,117 @@ ] }, { - "cluster_id": "cluster-46894-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #46894.", + "cluster_id": "cluster-46658-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", "status": "open", - "confidence": 0.8, - "canonical_issue_number": 46894, - "canonical_pr_number": 47090, - "issue_numbers": [ - 46894 - ], + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46658, + "issue_numbers": [], "pr_numbers": [ - 46958, - 47090 + 46658, + 46836 ], "evidence_types": [ - "closing_reference", - "shared_issue_target", "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46958, - "right_pr_number": 47090, - "code_similarity": 0.397, - "size_similarity": 0.257, - "file_overlap": 0.273, - "area_overlap": 0.598, - "patch_similarity": 0.362, + "left_pr_number": 46658, + "right_pr_number": 46836, + "code_similarity": 0.206, + "size_similarity": 0.71, + "file_overlap": 0.059, + "area_overlap": 0.0, + "patch_similarity": 0.308, "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" + "src/transformers/conversion_mapping.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" ], "shared_file_areas": [ { - "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "filename": "src/transformers/conversion_mapping.py", "left_ranges": [ [ - 22, - 28 - ], - [ - 638, - 645 + 739, + 830 ] ], "right_ranges": [ [ - 637, - 649 + 127, + 133 + ], + [ + 1495, + 1545 ] ] }, { - "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "filename": "src/transformers/models/auto/auto_mappings.py", "left_ranges": [ [ - 16, - 22 + 362, + 368 ], [ - 501, - 508 + 591, + 599 + ], + [ + 798, + 804 + ], + [ + 884, + 891 + ], + [ + 1084, + 1090 + ], + [ + 1207, + 1213 ] ], "right_ranges": [ [ - 500, - 512 + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 ] ] }, { - "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "filename": "src/transformers/models/auto/modeling_auto.py", "left_ranges": [ [ - 15, - 21 - ], - [ - 180, - 221 + 1085, + 1091 ] ], "right_ranges": [ [ - 175, - 224 + 509, + 515 + ], + [ + 1119, + 1125 ] ] } @@ -3144,6 +4252,65 @@ } ] }, + { + "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.", @@ -3203,60 +4370,105 @@ ] }, { - "cluster_id": "cluster-46836-2", - "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "cluster_id": "cluster-47173-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 46836, + "canonical_pr_number": 47449, "issue_numbers": [], "pr_numbers": [ - 46836, - 47387 + 47173, + 47292, + 47449 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46836, - "right_pr_number": 47387, - "code_similarity": 0.022, - "size_similarity": 0.049, + "left_pr_number": 47173, + "right_pr_number": 47292, + "code_similarity": 0.097, + "size_similarity": 0.463, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.027, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47173, + "right_pr_number": 47449, + "code_similarity": 0.1, + "size_similarity": 0.47, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47292, + "right_pr_number": 47449, + "code_similarity": 0.201, + "size_similarity": 0.984, "file_overlap": 0.0, "area_overlap": 0.0, - "patch_similarity": 0.082, + "patch_similarity": 0.029, "shared_filenames": [], "shared_file_areas": [] } ] }, { - "cluster_id": "cluster-46509-2", + "cluster_id": "cluster-46912-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", "status": "open", "confidence": 0.5, "canonical_issue_number": null, - "canonical_pr_number": 46509, + "canonical_pr_number": 47175, "issue_numbers": [], "pr_numbers": [ - 46509, - 47163 + 46912, + 47175 ], "evidence_types": [ "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46509, - "right_pr_number": 47163, - "code_similarity": 0.006, - "size_similarity": 0.012, - "file_overlap": 0.0, + "left_pr_number": 46912, + "right_pr_number": 47175, + "code_similarity": 0.009, + "size_similarity": 0.005, + "file_overlap": 0.022, "area_overlap": 0.0, - "patch_similarity": 0.024, - "shared_filenames": [], - "shared_file_areas": [] + "patch_similarity": 0.01, + "shared_filenames": [ + "tests/models/olmo3/test_modeling_olmo3.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/olmo3/test_modeling_olmo3.py", + "left_ranges": [ + [ + 99, + 104 + ] + ], + "right_ranges": [ + [ + 213, + 227 + ], + [ + 312, + 318 + ] + ] + } + ] } ] }, @@ -3538,12 +4750,12 @@ ], "duplicate_issues": [ { - "cluster_id": "cluster-44843-4", - "canonical_issue_number": 45081, + "cluster_id": "cluster-46894-4", + "canonical_issue_number": 46894, "duplicate_issue_numbers": [ - 44843 + 46897 ], - "reason": "Issues in cluster-44843-4 are treated as duplicates because they share closing_reference, shared_issue_target, soft_similarity evidence." + "reason": "Issues in cluster-46894-4 are treated as duplicates because they share closing_reference, soft_similarity evidence." }, { "cluster_id": "cluster-47138-4", @@ -3552,6 +4764,14 @@ 47139 ], "reason": "Issues in cluster-47138-4 are treated as duplicates because they share closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47301-4", + "canonical_issue_number": 47333, + "duplicate_issue_numbers": [ + 47332 + ], + "reason": "Issues in cluster-47301-4 are treated as duplicates because they share closing_reference, soft_similarity evidence." } ], "duplicate_prs": [ @@ -3573,6 +4793,18 @@ "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-43451-5", + "canonical_pr_number": 43451, + "duplicate_pr_numbers": [ + 45157, + 46509, + 47163, + 47306 + ], + "target_issue_number": null, + "reason": "PRs in cluster-43451-5 are treated as duplicates because they share soft_similarity evidence." + }, { "cluster_id": "cluster-43698-3", "canonical_pr_number": 43779, @@ -3609,13 +4841,13 @@ "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-44843-4", + "cluster_id": "cluster-45081-3", "canonical_pr_number": 45317, "duplicate_pr_numbers": [ 45086 ], "target_issue_number": 45081, - "reason": "PRs in cluster-44843-4 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + "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", @@ -3653,15 +4885,6 @@ "target_issue_number": 46305, "reason": "PRs in cluster-46305-3 are treated as duplicates because they converge on issue #46305 with closing_reference, shared_issue_target, soft_similarity evidence." }, - { - "cluster_id": "cluster-46509-2", - "canonical_pr_number": 46509, - "duplicate_pr_numbers": [ - 47163 - ], - "target_issue_number": null, - "reason": "PRs in cluster-46509-2 are treated as duplicates because they share soft_similarity evidence." - }, { "cluster_id": "cluster-46614-3", "canonical_pr_number": 46687, @@ -3672,13 +4895,13 @@ "reason": "PRs in cluster-46614-3 are treated as duplicates because they converge on issue #46614 with closing_reference, shared_issue_target, soft_similarity evidence." }, { - "cluster_id": "cluster-46836-2", - "canonical_pr_number": 46836, + "cluster_id": "cluster-46658-2", + "canonical_pr_number": 46658, "duplicate_pr_numbers": [ - 47387 + 46836 ], "target_issue_number": null, - "reason": "PRs in cluster-46836-2 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-46658-2 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-46869-3", @@ -3690,19 +4913,29 @@ "reason": "PRs in cluster-46869-3 are treated as duplicates because they converge on issue #46869 with closing_reference, shared_issue_target evidence." }, { - "cluster_id": "cluster-46894-3", + "cluster_id": "cluster-46894-4", "canonical_pr_number": 47090, "duplicate_pr_numbers": [ 46958 ], "target_issue_number": 46894, - "reason": "PRs in cluster-46894-3 are treated as duplicates because they converge on issue #46894 with closing_reference, shared_issue_target, soft_similarity evidence." + "reason": "PRs in cluster-46894-4 are treated as duplicates because they converge on issue #46894 with closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46912-2", + "canonical_pr_number": 47175, + "duplicate_pr_numbers": [ + 46912 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46912-2 are treated as duplicates because they share soft_similarity evidence." }, { - "cluster_id": "cluster-47122-11", + "cluster_id": "cluster-47122-13", "canonical_pr_number": 47122, "duplicate_pr_numbers": [ 47123, + 47124, 47224, 47266, 47280, @@ -3711,10 +4944,11 @@ 47310, 47380, 47443, - 47444 + 47444, + 47445 ], "target_issue_number": null, - "reason": "PRs in cluster-47122-11 are treated as duplicates because they share soft_similarity evidence." + "reason": "PRs in cluster-47122-13 are treated as duplicates because they share soft_similarity evidence." }, { "cluster_id": "cluster-47138-4", @@ -3724,18 +4958,37 @@ ], "target_issue_number": 47138, "reason": "PRs in cluster-47138-4 are treated as duplicates because they converge on issue #47138 with closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47173-3", + "canonical_pr_number": 47449, + "duplicate_pr_numbers": [ + 47173, + 47292 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47173-3 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47301-4", + "canonical_pr_number": 47336, + "duplicate_pr_numbers": [ + 47301 + ], + "target_issue_number": 47333, + "reason": "PRs in cluster-47301-4 are treated as duplicates because they converge on issue #47333 with closing_reference, soft_similarity evidence." } ], "best_issue": { - "cluster_id": "cluster-47138-4", - "issue_number": 47138, - "reason": "Issue #47138 is the strongest global issue candidate because it is open, belongs to a cluster with 4 artifacts, and carries 1 discussion comments plus 3 inbound references.", - "score": 145.0 + "cluster_id": "cluster-47301-4", + "issue_number": 47333, + "reason": "Issue #47333 is the strongest global issue candidate because it is open, belongs to a cluster with 4 artifacts, and carries 5 discussion comments plus 4 inbound references.", + "score": 154.0 }, "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 + "cluster_id": "cluster-47122-13", + "pr_number": 47122, + "reason": "PR #47122 is the strongest global PR candidate because it is open, belongs to a cluster with 13 artifacts, links to 0 issues, and carries 8 review/discussion events.", + "score": 265.974 } } diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 53de08fde61f116214aea66b300bd0aba6a8fd7d..b38a8f157556a1d2ad38dd83c26d6a82c4f95b2a 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": "20260721T180022Z", - "generated_at": "2026-07-21T18:20:54Z", + "snapshot_id": "20260722T000038Z", + "generated_at": "2026-07-22T00:15:44Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -85,8 +85,8 @@ "issue:35545|issue:41720", "issue:45901|issue:46015", "issue:35545|issue:36010", - "issue:44509|issue:44512", "issue:44918|issue:46165", + "issue:44509|issue:44512", "issue:46165|issue:46869", "issue:44805|issue:45823", "issue:43638|issue:44805", @@ -131,376 +131,406 @@ "issue:43039|issue:44918", "issue:41762|issue:43638" ], - "prepared_review_unit_hash": "cf2069759dc61095d62a46725beaffe0e4146f1967b6cac069e771ed3ce4a53c", + "prepared_review_unit_hash": "0bb96df5b444835f6f3bd8214c8d2c1027165b24e2cae5705a30365804380bc5", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 62, - "serialized_chars": 46353, - "estimated_input_tokens": 11589, - "estimated_eval_tokens": 23434 + "serialized_chars": 46364, + "estimated_input_tokens": 11591, + "estimated_eval_tokens": 23438 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a mixed bag of unrelated issues with a few real duplicate subclusters. The clearest duplicates are the DeepSpeed ZeRO-3 index-out-of-bounds reports, the torch 2.4.1 compatibility reports, and the Qwen3.5MoE rotary-embedding materialization/offload regressions; the two docs-v5 cleanup issues also look like the same documentation drift.", - "confidence": 0.81, - "canonical_issue_reason": "Issue 41762 is the strongest anchor for the genuine duplicate subcluster: it has the clearest concrete DeepSpeed ZeRO-3 load failure (`index 0 is out of bounds for dimension 0 with size 0`) and substantial discussion, making it the best representative of that bug.", + "summary": "This cluster is mostly heterogeneous. The only clear duplicate subclusters are the Zero-3/BERT index-error reports, the v5 docs cleanup reports, and a couple of closely related torch/version or Qwen3.5-MoE embedding issues.", + "confidence": 0.78, + "canonical_issue_reason": "issue:43596 is the clearest representative of the strongest duplicate pair (with 43638) and has the most concrete, reproducible Zero-3/BERT crash signature.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 41762 is the best overall issue in this set because it is the most clearly scoped, reproducible, and discussed report of the zero-sized tensor/index failure path; it also matches the broader ZeRO-3 duplicate cluster better than the more incidental reports.", + "best_issue_reason": "issue:43596 is the most actionable standalone bug report in the set: precise error, clear code path, and a near-identical duplicate companion.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different failures: Qwen3.5 config attribute error vs chat/completions streaming inference on Qwen3.5-0.8B." + "reason": "Different failure modes and code paths: config attribute error vs streaming chat/completions inference." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Different problems: AMD GPU failure for GPT-OSS-20B vs MPS histogram kernel not implemented for Int." + "reason": "Different subsystems and bugs: AMD GPU failure vs MPS histogram op limitation." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different code paths: image-preprocessor dtype mismatch vs GPT-OSS on AMD GPUs." + "reason": "Different models and root causes: image-preprocessor dtype mismatch vs AMD GPU support issue." }, { "left": "issue:39921", "right": "issue:44918", "accept": false, - "reason": "Both involve model/token handling, but the reported bugs are different: Gemma3N special-token projection vs Qwen3.5 input embeddings with TRL SFT." + "reason": "Different models and errors: special-token projection issue vs TRL SFT embedding unpacking." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Autocast dtype mismatch is not the same as the AMD GPU device-map failure." + "reason": "Both are dtype/GPU-related, but they describe different models and concrete failures." }, { "left": "issue:39921", "right": "issue:43039", "accept": false, - "reason": "Unrelated issues: special-token projection error vs Liger Kernel cross-entropy being called." + "reason": "Gemma3N token projection bug is unrelated to Liger Kernel cross-entropy dispatch." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Different models and failures; no shared underlying bug." + "reason": "Unrelated model-specific bugs with no shared concrete code path." }, { "left": "issue:39921", "right": "issue:45685", "accept": false, - "reason": "Gemma3N token projection and MPS Int histogram error are unrelated." + "reason": "Different model families and different failures; no shared underlying change." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "GPT-OSS AMD GPU failure and GraniteMoEHybrid invalid-method call are distinct." + "reason": "Different model/runtime failures; both MoE-ish at a high level but not the same bug." }, { "left": "issue:35545", "right": "issue:39921", "accept": false, - "reason": "ModernBERT ONNX export error is unrelated to Gemma3N special-token projection." + "reason": "Unrelated issues: ONNX export vs special-token projection handling." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "DeepSpeed collective mismatch and ONNX export are unrelated failures." + "reason": "DeepSpeed collective mismatch and ONNX export error are different code paths." }, { "left": "issue:39921", "right": "issue:44805", "accept": false, - "reason": "Special-token projection error and mask/tensor shape mismatch are different bugs." + "reason": "Different bug classes: token projection vs tensor mask shape mismatch." }, { "left": "issue:39921", "right": "issue:43716", "accept": false, - "reason": "Gemma3N special-token issue does not match the image dtype mismatch bug." + "reason": "Different model-specific failures with no shared code path." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "ONNX export failure is unrelated to Qwen3Moe FineGrainedFP8Config failure." + "reason": "ONNX export and FineGrainedFP8 model execution are unrelated." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "ModernBERT ONNX export and Qwen3 auto-device-map cuda assert are unrelated." + "reason": "Different models and failures; no concrete overlap beyond general inference tooling." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "Different topics: table-question-answering crash vs outdated RAG docs." + "reason": "One is a table-QA crash, the other is outdated RAG documentation." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export issue is unrelated to GenerationMixin import failure." - }, - { - "left": "issue:44509", - "right": "issue:44512", - "accept": true, - "reason": "Both are the same kind of v5 documentation cleanup: stale docs still mention removed features/commands." + "reason": "ONNX export bug is unrelated to the GenerationMixin import regression." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "Different issues: TRL SFT embedding unpacking vs training resume with trust_remote_code." + "reason": "Different training/resume issues with no shared concrete bug." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": true, + "reason": "Both are v5 docs cleanup reports about stale removed commands/tasks; same documentation-maintenance change." }, { "left": "issue:46165", "right": "issue:46869", "accept": false, - "reason": "Resume/trust_remote_code problem is unrelated to YOLO multi-GPU device-map auto failure." + "reason": "Different problems: resume-with-remote-code vs multi-GPU YOLO device-map failure." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Mask-shape mismatch and Gemma4 PLE device mismatch are not the same underlying bug." + "reason": "Mask-shape mismatch and Gemma4 device mismatch are distinct runtime errors." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Both mention errors during loading/training, but the specific failures differ: ZeRO-3 index OOB vs mask/tensor shape mismatch." - }, - { - "left": "issue:43638", - "right": "issue:46165", - "accept": false, - "reason": "DeepSpeed ZeRO-3 index OOB is unrelated to resume/trust_remote_code handling." + "reason": "Different error signatures and different model paths." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Different concrete failures: ZeRO-3 index OOB vs mask shape mismatch." + "reason": "Zero-3/BERT index error is unrelated to the mask-indexing shape mismatch." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Mask/index shape mismatch is unrelated to video feature/token count mismatch." + "reason": "Different modalities and failure sites; not the same underlying bug." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Model resume issue and Nemotron example breakage are unrelated." + "reason": "Unrelated issues with no shared code path." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Mask-shape indexing bug vs GPT-OSS MoE TP failure are different problems." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video StopIteration in Qwen3-vl is unrelated to resuming training with remote code." + "reason": "Different feature areas: Qwen3-VL chat-template video error vs training resume issue." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both involve DeepSpeed, but one is a collective mismatch and the other is a zero-sized tensor index error." + "reason": "Both involve DeepSpeed, but the concrete failures and models differ." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same failure mode: DeepSpeed ZeRO-3 initialization/loading hits `index 0 is out of bounds for dimension 0 with size 0` on BERT models." + "reason": "Same underlying bug: DeepSpeed ZeRO-3/BERT loading hits the same index-0-out-of-bounds crash." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy call and GPT-OSS TP failure are different bugs." + "reason": "Liger Kernel cross-entropy dispatch is unrelated to GPT-OSS MoE tensor-parallel failure." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Incorrect detokenization and LoRA merge/save/load output mismatch are not the same bug." + "reason": "Different tasks: detokenization regression vs LoRA merge/save/load inconsistency." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime failure and LoRA save/reload inconsistency are unrelated." + "reason": "Different model/runtime failures; no shared code path." }, { "left": "issue:34928", "right": "issue:35308", "accept": false, - "reason": "Activation-checkpoint/FSDP recompute size mismatch and iterable-dataset length crash are different training failures." + "reason": "Both are multi-GPU/training issues, but the symptoms and causes are different." }, { "left": "issue:44841", "right": "issue:46661", "accept": false, - "reason": "Processor failure for Voxtral and remote-code `config_class` AttributeError are unrelated." + "reason": "Processor failure and remote-code model_class/config_class error are unrelated." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Different DeepSpeed errors on different code paths; not the same underlying bug." + "reason": "Both mention DeepSpeed, but the actual failures are different and model-specific." }, { "left": "issue:43716", "right": "issue:46869", "accept": false, - "reason": "Image-preprocessor dtype mismatch and YOLO multi-GPU device-map failure are unrelated." + "reason": "Different models and different hardware/device-map failures." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Different shape mismatch contexts and triggers; not the same bug." + "reason": "Activation-checkpointing/FSDP size mismatch is not the same as a mask indexing error." }, { "left": "issue:45685", "right": "issue:46869", "accept": false, - "reason": "MPS histogram kernel limitation is unrelated to YOLO multi-GPU device-map auto failure." + "reason": "MPS histogram limitation is unrelated to multi-GPU device-map auto behavior." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "Different training bugs: recomputed-tensor mismatch vs custom trainer signature change." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": true, + "reason": "Same Qwen3.5-MoE rotary-embedding area with offload/materialization bugs; likely the same underlying fix." }, { "left": "issue:43638", "right": "issue:45161", "accept": false, - "reason": "ZeRO-3 index OOB and GPT-OSS TP-only failure are different issues." + "reason": "Zero-3/BERT crash is unrelated to GPT-OSS MoE tensor-parallel failure." }, { - "left": "issue:43975", - "right": "issue:44457", + "left": "issue:34928", + "right": "issue:39921", + "accept": false, + "reason": "FSDP/activation-checkpointing issue is unrelated to Gemma3N special-token projection." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different model families and different runtime paths." + }, + { + "left": "issue:43716", + "right": "issue:45507", "accept": false, - "reason": "Detokenization bug and LoRA weight merge inconsistency are different regressions." + "reason": "Different model-specific failures with no shared concrete bug." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Qwen-3-8B-NVFP4 failure and GPT-OSS AMD GPU failure are not the same bug." + "reason": "Different model/runtime issues; no same code path." }, { - "left": "issue:34928", - "right": "issue:41720", + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Expensive reverse-embedding lookup and video chat-template token mismatch are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45081", "accept": false, - "reason": "FSDP/activation-checkpoint mismatch is unrelated to Qwen3 auto-device-map CUDA assert." + "reason": "Different bugs: Gemma3 text config regression vs Mistral regex patch crash." }, { "left": "issue:45800", "right": "issue:46515", "accept": true, - "reason": "Both report the same torch 2.4.1 compatibility breakage." + "reason": "Both report the same torch 2.4.1 incompatibility regression; one is just more specific." }, { - "left": "issue:43716", - "right": "issue:45507", + "left": "issue:34928", + "right": "issue:41720", "accept": false, - "reason": "Image dtype mismatch and GraniteMoEHybrid invalid-method call are unrelated." + "reason": "Different model/runtime failures with no shared code path." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Device mismatch in Gemma4 PLE and expensive reverse embedding lookup are different forward-path problems." + "reason": "Device-map forward mismatch and reverse-embedding lookup are different bugs." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Generic torch compatibility issue and Qwen-3-8B-NVFP4 runtime error are unrelated." + "reason": "Generic torch incompatibility and Qwen3-8B NVFP4 runtime error are not the same issue." }, { "left": "issue:43828", "right": "issue:45161", "accept": false, - "reason": "Autocast dtype mismatch is not the same as the GPT-OSS TP-only issue." + "reason": "Different model families and different dtype/device-parallel failures." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Torch dynamo grouped_mm tracing failure is unrelated to image preprocessor dtype mismatch." + "reason": "TorchDynamo grouped_mm tracing error is unrelated to the Mistral image dtype mismatch." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Different failures and different models; no duplicate evidence." + "reason": "Different model/runtime bugs with no shared underlying change." }, { "left": "issue:43716", "right": "issue:45685", "accept": false, - "reason": "Mistral image dtype mismatch and MPS histogram Int error are unrelated." + "reason": "Different models and backends; not the same bug." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid and MPS histogram kernel limitation are unrelated." + "reason": "Different code paths and error types; only broad MoE/backend overlap." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Qwen-3-8B-NVFP4 runtime issue and TRL SFT embedding unpacking are different problems." + "reason": "Different model-specific runtime failures." }, { "left": "issue:43716", "right": "issue:44292", "accept": false, - "reason": "Different models and failure modes; not the same underlying bug." + "reason": "Different models and errors; no concrete duplicate signal." }, { "left": "issue:44918", "right": "issue:45161", "accept": false, - "reason": "TRL SFT embedding issue and GPT-OSS TP-only issue are unrelated." + "reason": "TRl SFT embedding unpacking and GPT-OSS TP failure are unrelated." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger Kernel cross-entropy call and TRL SFT embedding unpacking are different bugs." + "reason": "Liger Kernel cross-entropy issue is unrelated to Qwen3.5 embedding unpacking." }, { "left": "issue:41762", "right": "issue:43638", - "accept": true, - "reason": "Same concrete DeepSpeed ZeRO-3 bug: loading/training with zero-sized tensors leads to `index 0 is out of bounds for dimension 0 with size 0`." + "accept": false, + "reason": "Gemma3 ZeRO-3 loading failure and BERT ZeRO-3 loading failure are similar in shape, but not the same concrete bug." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The accepted soft pairs are well supported by the titles/discussion context, especially the DeepSpeed ZeRO-3 duplicates, the torch 2.4.1 compatibility pair, and the docs-v5 cleanup pair. I don\u2019t see a clear overstatement that would require rejection." + "feedback": "The summary is well grounded in the packet: the cluster is mostly heterogeneous, and the identified duplicate subclusters are the ones with strong title/body similarity (Zero-3/BERT index errors, v5 docs cleanup, torch 2.4.1 incompatibility, and the Qwen3.5-MoE rotary-embedding pair). The soft-edge rejections are generally conservative and tied to distinct titles/error modes. No clear overstatement stands out." }, "evaluator_used": true, "retried": false, @@ -553,8 +583,6 @@ "issue:45020", "issue:45070", "issue:45237", - "issue:45362", - "issue:45464", "issue:45800", "issue:45812", "issue:45823", @@ -565,6 +593,7 @@ "issue:46304", "issue:46402", "issue:46772", + "issue:46869", "issue:46901", "issue:47228", "issue:47394", @@ -594,270 +623,288 @@ "issue:35308|issue:42617", "issue:45812|issue:45920", "issue:43872|issue:46304", - "issue:46901|issue:47394", "issue:44292|issue:46015", + "issue:46901|issue:47394", "issue:39290|issue:43531", "issue:43866|issue:44863", "issue:41093|issue:43638", "issue:43792|issue:45901", - "issue:43854|issue:43866", "issue:43425|issue:45070", + "issue:43854|issue:43866", "issue:43541|issue:47228", - "issue:43975|issue:45901", "issue:43792|issue:45070", + "issue:43975|issue:45901", "issue:43596|issue:45800", - "issue:45362|issue:45464" + "issue:43541|issue:46869" ], - "prepared_review_unit_hash": "14086974c4bba7b8d56d247fe3710494e9ff249ca2de5d29e5ff3b73968f7819", + "prepared_review_unit_hash": "5fa8fb3c0efd610dd3510ad47a55d951489cafb5a056ed258823d2a97636d60d", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 36, - "serialized_chars": 38336, - "estimated_input_tokens": 9584, - "estimated_eval_tokens": 19424 + "serialized_chars": 37776, + "estimated_input_tokens": 9444, + "estimated_eval_tokens": 19144 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly a set of unrelated issue reports around different model-specific regressions, docs breakages, and compatibility problems. I only see two strong duplicate pairs: the ZeRO-3 zero-length index error during model loading, and the v5 AutoTokenizer wrong-token regression across model families. No pull requests were provided.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45920 is the best canonical issue because it describes the broader, cross-model AutoTokenizer regression in v5 and is more representative than the narrower Granite-only report.", - "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the strongest global issue representative: broad impact, clear regression framing, and it captures the shared tokenizer bug affecting multiple model families.", - "best_pr_reason": null, + "summary": "This set is mostly heterogeneous; only two soft duplicate pairs look convincing: a DeepSpeed ZeRO-3 zero-sized-index loading failure and an AutoTokenizer v5 regression producing wrong token IDs. No PRs are present.", + "confidence": 0.58, + "canonical_issue_reason": "issue:45920 is the broadest and most representative standalone report of the tokenizer regression, covering multiple model families rather than a single model-specific symptom.", + "canonical_pr_reason": "No PRs in the cluster.", + "best_issue_reason": "issue:45920 has the clearest cross-model scope and discussion for the tokenizer ID regression, making it the best issue representative among the accepted duplicate subgroup.", + "best_pr_reason": "No PRs available.", "soft_edge_verdicts": [ { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs problems, but they refer to different broken references and different pages; not the same underlying change." + "reason": "Both are docs cleanup, but they refer to different removed behaviors/commands and do not describe the same concrete change." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Different failures: Liger Kernel cross_entropy usage vs autocast dtype mismatch." + "reason": "Different bugs: Liger cross-entropy routing vs an autocast dtype mismatch in Phi-tiny-MoE." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "Unrelated training/runtime issues: ZeRO-3 collective mismatch versus a CustomTrainer signature error." + "reason": "Different failure modes: DeepSpeed collective mismatch vs a trainer signature TypeError." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "Different domains entirely: AMD GPU runtime failure versus outdated RAG documentation." + "reason": "Unrelated: AMD GPU runtime failure vs outdated RAG docs." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Cross-entropy/Liger Kernel issue is unrelated to the mask/tensor shape mismatch." + "reason": "Different code paths: Liger kernel cross-entropy vs a mask/tensor shape mismatch." }, { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Both are compatibility regressions, but they involve different constraints and code paths (torch version mismatch vs USE_DISTRIBUTED=0 support)." + "reason": "Both mention compatibility, but these are distinct regressions in different environments and code paths." }, { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "Different model-loading failure versus a PyTorch distributed build/config regression." + "reason": "Unrelated model-loading/vLLM failure vs a USE_DISTRIBUTED=0 build/runtime regression." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Both are model-specific runtime failures, but the vLLM Gemma3 config error and Qwen3 auto device-map cuda assert are not the same bug." + "reason": "Different problems: missing Gemma3 config attribute in vLLM vs Qwen3 auto device-mapping CUDA assert." }, { "left": "issue:43866", "right": "issue:46402", "accept": false, - "reason": "Checkpoint corruption/loading failure is distinct from missing text encoder weights in Sam3Model." + "reason": "Different model-specific load failures; no shared underlying bug is evident." }, { "left": "issue:44841", "right": "issue:47448", "accept": false, - "reason": "Different model families and different failures: Voxtral processor bug versus Gemma4 architecture recognition." + "reason": "Both are model-architecture issues, but they concern different models and different loading/processor failures." }, { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Float8 storage lookup error is unrelated to the distributed-support regression." + "reason": "Different regressions: missing Float8 storage type vs distributed build support breakage." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "Different tasks and failure modes: AMD GPU support versus table-question-answering crash." + "reason": "Unrelated runtime failure vs table-question-answering crash." }, { "left": "issue:41762", "right": "issue:43596", "accept": true, - "reason": "Both report the same ZeRO-3 loading/init bug: index 0 out of bounds from an empty tensor/parameter shard, just on different models." + "reason": "Same underlying DeepSpeed ZeRO-3 zero-sized-index failure during model loading; only the affected model differs." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs issues, but they cover different APIs and different removed/changed references." + "reason": "Separate documentation issues referencing different removed APIs/commands." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Different model-specific runtime paths: Voxtral processor failure versus Gemma4 PLE device placement mismatch." + "reason": "Different model-specific failures: processor loading vs device-mapping forward mismatch." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Device-map forward mismatch and chat-template video token mismatch are separate multimodal bugs." + "reason": "Both are multimodal issues, but they target different mechanisms (device mapping vs chat-template video token alignment)." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Detokenization regression is unrelated to outdated RAG examples." + "reason": "Different topics: detokenization bug vs outdated RAG examples." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Different compatibility regressions: remote_code breakage versus PyTorch distributed support breakage." + "reason": "Remote_code model regressions are not the same as the USE_DISTRIBUTED=0 compatibility break." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "NemotronH checkpoint loading is unrelated to RAG documentation." + "reason": "Checkpoint-loading issue vs documentation drift; unrelated." }, { "left": "issue:43866", "right": "issue:46772", "accept": false, - "reason": "Checkpoint corruption is not the same as ModelOpt/NVFP4 compatibility for DiffusionGemma." + "reason": "Different model families and different compatibility concerns." }, { "left": "issue:35308", "right": "issue:42617", "accept": false, - "reason": "Both involve distributed training, but one is IterableDataset length imbalance and the other is a 3d_parallel.py run failure." + "reason": "Both involve training/distributed execution, but the concrete bugs are different." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "These describe the same tokenizer regression in v5; 45920 is the broader report, while 45812 is a narrower Granite subset." + "reason": "Same tokenizer regression: wrong token IDs across models, with 45920 being the broader multi-model report." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes Int8Params constructor incompatibility is a different dependency break than the distributed-support regression." + "reason": "bitsandbytes parameter incompatibility is unrelated to distributed build/runtime support." }, { - "left": "issue:46901", - "right": "issue:47394", + "left": "issue:44292", + "right": "issue:46015", "accept": false, - "reason": "Both are training-loss shifting bugs, but they are model-specific and describe different concrete defects." + "reason": "Unrelated NVFP4 model runtime issue vs documentation cleanup." }, { - "left": "issue:44292", - "right": "issue:46015", + "left": "issue:46901", + "right": "issue:47394", "accept": false, - "reason": "Qwen-3 NVFP4 runtime failure is unrelated to RAG docs." + "reason": "Both are loss-shift bugs, but they are model-specific and describe different failure details; not clearly the same underlying bug." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both mention sliding window, but the model families and reported failures differ too much to treat as the same concrete bug." + "reason": "Different Gemma/Qwen model issues with different symptoms and code paths." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Corrupted checkpoint and inability to load checkpoints are not the same reported issue." + "reason": "Different model-loading failures with no clear shared root cause." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Mask/tensor shape mismatch is a different error from ZeRO-3 zero-length index out-of-bounds." + "reason": "Both mention index errors, but they appear to be separate shape/ZeRO-3 problems rather than the same bug." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper loading failure is unrelated to table-question-answering crashes." + "reason": "Different tasks and different runtime failures." }, { - "left": "issue:43854", - "right": "issue:43866", + "left": "issue:43425", + "right": "issue:45070", "accept": false, - "reason": "GLM-4.7-Flash load issues are not the same as an Ovis2 checkpoint corruption report." + "reason": "Torch-version incompatibility is not the same as the PretrainedConfig pydantic regression." }, { - "left": "issue:43425", - "right": "issue:45070", + "left": "issue:43854", + "right": "issue:43866", "accept": false, - "reason": "Torch version incompatibility and a pydantic PretrainedConfig regression are different problems." + "reason": "Unable-to-load test failure vs checkpoint corruption; not the same underlying issue." }, { "left": "issue:43541", "right": "issue:47228", "accept": false, - "reason": "Both are compilation/tracing regressions, but they affect different models and different ops/code paths." + "reason": "Both concern model compilation, but they are different models and different regressions." }, { - "left": "issue:43975", - "right": "issue:45901", + "left": "issue:43792", + "right": "issue:45070", "accept": false, - "reason": "Detokenization regression is unrelated to table QA crashes." + "reason": "Unrelated model loading issue vs pydantic config regression." }, { - "left": "issue:43792", - "right": "issue:45070", + "left": "issue:43975", + "right": "issue:45901", "accept": false, - "reason": "Whisper model loading is unrelated to the PretrainedConfig pydantic breakage." + "reason": "Different bugs affecting different functionality; no shared code-path issue is apparent." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "ZeRO-3 model init failure is unrelated to the torch 2.4.1 / transformers 5.8.0 compatibility issue." + "reason": "ZeRO-3 zero-sized tensor indexing is unrelated to torch/transformers version incompatibility." }, { - "left": "issue:45362", - "right": "issue:45464", + "left": "issue:43541", + "right": "issue:46869", + "accept": false, + "reason": "Different model/runtime regressions; no concrete shared bug." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures and different affected artifacts." + }, + { + "left": "issue:43866", + "right": "issue:46772", "accept": false, - "reason": "Both are Qwen3.5 chat-related, but one is transformers chat crashing and the other is streaming chat/completions API failure; not enough evidence for the same bug." + "reason": "Different compatibility/load issues with no common code-path evidence." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same tokenizer-ID regression reported on different model sets; 45920 is the broader manifestation." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative overall. The two accepted duplicate pairs are well supported by the issue titles: the ZeRO-3 zero-length index failures and the v5 AutoTokenizer regression. I do not see evidence of overstatement or an unsafe merge of unrelated pairs." + "feedback": "The summary is grounded in the packet: there are no PRs, and the two accepted soft pairs are the only ones that look plausibly duplicative from the issue titles. The canonical issue choice is conservative and supported by the broader multi-model scope of issue:45920. No clear overstatement of evidence." }, "evaluator_used": true, "retried": false, @@ -884,7 +931,6 @@ "issue:40990", "issue:41093", "issue:41762", - "issue:42915", "issue:43054", "issue:43296", "issue:43366", @@ -911,6 +957,7 @@ "issue:45020", "issue:45070", "issue:45084", + "issue:45362", "issue:45399", "issue:45464", "issue:45507", @@ -928,7 +975,7 @@ "issue:47228" ], "soft_pairs": [ - "issue:43541|issue:46869", + "issue:45362|issue:45464", "issue:44292|issue:45901", "issue:39290|issue:45823", "issue:36010|issue:39290", @@ -958,280 +1005,278 @@ "issue:43716|issue:46294", "issue:44479|issue:46294", "issue:41093|issue:43596", - "issue:43054|issue:43994", "issue:39921|issue:46869", + "issue:43054|issue:43994", "issue:45759|issue:45902", "issue:46640|issue:46661", "issue:43596|issue:46165", "issue:43716|issue:44918", "issue:43645|issue:45399", "issue:39921|issue:45507", - "issue:40990|issue:44960", - "issue:42915|issue:45399", - "issue:43404|issue:45005" + "issue:40990|issue:44960" ], - "prepared_review_unit_hash": "13dbbde31cc00681ffcdfb7b7262702a80074b6ad0021ed502554b9ab6c6169f", + "prepared_review_unit_hash": "c3a3b8572b71370165ddafba102cf890cc7476a4b7ee0e62bc59ae00bf3b9062", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 41, - "serialized_chars": 39886, - "estimated_input_tokens": 9972, - "estimated_eval_tokens": 20200 + "soft_pair_count": 39, + "serialized_chars": 39282, + "estimated_input_tokens": 9821, + "estimated_eval_tokens": 19898 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a noisy mix of unrelated model, config, quantization, multimodal, and runtime regressions. None of the soft pairs look like the same underlying bug or a mergeable duplicate PR, so there is no strong canonical artifact.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This cluster is heterogeneous: it mixes unrelated Transformers bug reports across model loading, multimodal processors, quantization, vLLM/DeepSpeed, and backend-specific failures. I reject all soft edges; none look like the same concrete bug/change. If forced to name a representative issue, issue:45020 is the broadest regression-style report, but it is only a weak umbrella fit, not a true duplicate hub.", + "confidence": 0.88, + "canonical_issue_reason": "Issue:45020 is the most general regression-style report in the set (recent Transformers versions breaking remote_code models), so it is the best weak umbrella candidate among otherwise unrelated bugs.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue:45020 is the broadest and most reusable representative issue if a single anchor must be chosen, though the cluster is too mixed for a strong canonical duplicate.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43541", - "right": "issue:46869", + "left": "issue:45362", + "right": "issue:45464", "accept": false, - "reason": "Different failures: Mixtral torch-dynamo grouped_mm tracing vs YOLO multi-GPU device-map behavior." + "reason": "Same Qwen3.5 family, but different failures: chat/streaming crash vs chat template API failure. Not the same bug." }, { "left": "issue:44292", "right": "issue:45901", "accept": false, - "reason": "One is an `_is_hf_initialized` init_empty_weights TypeError; the other is a table-question-answering task crash." + "reason": "Unexpected _is_hf_initialized in init_empty_weights is unrelated to a table-question-answering crash." }, { "left": "issue:39290", "right": "issue:45823", "accept": false, - "reason": "Gemma3/vLLM missing `sliding_window_pattern` is unrelated to Gemma4 PLE device mismatch." + "reason": "Gemma3TextConfig sliding_window_pattern error is a different code path from Gemma4 PLE device mismatch." }, { "left": "issue:36010", "right": "issue:39290", "accept": false, - "reason": "Importing `GenerationMixin` and a Gemma3 config attribute error are separate issues." + "reason": "ImportError for GenerationMixin and Gemma3TextConfig attr missing are unrelated issues." }, { "left": "issue:45464", "right": "issue:45685", "accept": false, - "reason": "Qwen3.5 streaming API failure is not the same code path as the MPS `histogram_mps` Int error." + "reason": "Streaming chat failure on Qwen3.5 and an MPS histogram kernel error are unrelated." }, { "left": "issue:43541", "right": "issue:45685", "accept": false, - "reason": "No shared concrete bug; one is torch-dynamo grouped_mm, the other is MPS/moe histogram handling." + "reason": "Torch Dynamo grouped_mm tracing failure is not the same as the MPS Int histogram error." }, { "left": "issue:43541", "right": "issue:43828", "accept": false, - "reason": "Both mention runtime errors, but the affected model and root cause differ: grouped_mm tracing vs autocast dtype mismatch." + "reason": "Both involve runtime math issues, but grouped_mm tracing and autocast dtype mismatch are different bugs." }, { "left": "issue:44841", "right": "issue:45084", "accept": false, - "reason": "Processor failure for Voxtral is unrelated to the non-template-node compilation error." + "reason": "Voxtral processor failure and non-template-node compile error are different subsystems and symptoms." }, { "left": "issue:45507", "right": "issue:46869", "accept": false, - "reason": "Invalid method on GraniteMoEHybrid and YOLO multi-GPU failures are unrelated." + "reason": "GraniteMoEHybrid invalid method and YOLO multi-GPU device_map failure are unrelated." }, { "left": "issue:43404", "right": "issue:43866", "accept": false, - "reason": "Mistral3 tied-weights bug and Ovis2 checkpoint corruption are different underlying problems." + "reason": "lm_head tying bug in Mistral3 and a corrupted Ovis2 checkpoint are not the same issue." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Both are dtype/Int-related runtime errors, but they affect different models and mechanisms." + "reason": "Autocast dtype mismatch and MPS histogram_mps not implemented are unrelated backend bugs." }, { "left": "issue:45399", "right": "issue:46661", "accept": false, - "reason": "Flash-Attn2 fallback gating and remote-code `config_class` lookup are distinct breakages." + "reason": "flash-attn2 fallback logic and remote-code model_class/config_class handling are different code paths." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Missing `pad_token_idx` in StableLmConfig is not the same as the pydantic `PretrainedConfig` field regression." + "reason": "Both are 5.0 config regressions, but missing pad_token_idx and broken PretrainedConfig pydantic field are distinct bugs." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are IndexErrors, but one is a mask/tensor length mismatch and the other is a zero-sized tensor during Gemma3/ZeRO-3 loading." + "reason": "Both are IndexErrors, but the mask/tensor shape mismatch is not the same as the zero-size index error under ZeRO-3." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "Torch/Transformers version incompatibility is a broader environment issue, not the same as Gemma4 device placement mismatch." + "reason": "Torch/Transformers incompatibility and Gemma4 device placement mismatch are separate issues." }, { "left": "issue:43872", "right": "issue:47228", "accept": false, - "reason": "bitsandbytes `_is_hf_initialized` TypeError is unrelated to the Sam3 bf16 compilation regression." + "reason": "bitsandbytes _is_hf_initialized incompatibility and Sam3 bf16 compilation regression are unrelated." }, { "left": "issue:43716", "right": "issue:47228", "accept": false, - "reason": "Image-preprocessor dtype mismatch and Sam3 bf16 compilation are different model-path failures." + "reason": "Mistral-3 image preprocessor dtype mismatch and Sam3 bf16 compilation regression do not share the same failure path." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Mask shape mismatch and Qwen3.5 embedding unpacking fail in different code paths." + "reason": "Similar index/shape wording, but the generic mask mismatch and Qwen3.5 embedding unpacking under TRL are different bugs." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Phi-tiny-MoE autocast dtype mismatch is not the same issue as Qwen3.5 streaming chat/completions failure." + "reason": "Autocast dtype mismatch is not the same as a Qwen3.5 streaming inference/API failure." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "PaddleOCR-VL load failure and GGUF gpt-oss support are different model-support problems." + "reason": "PaddleOCR-VL loading in vLLM and gpt-oss GGUF architecture support are unrelated." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3.5 streaming inference failure are unrelated." + "reason": "GGUF gpt-oss support and Qwen3.5 streaming chat failure are not the same bug." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "`init_empty_weights` unexpected-argument TypeError and pydantic config-field breakage are separate compatibility bugs." + "reason": "Unexpected _is_hf_initialized in init_empty_weights and a pydantic PretrainedConfig field break are different regressions." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Qwen3-MoE sliding_window handling is unrelated to the v5 pydantic `PretrainedConfig` regression." + "reason": "Qwen3-MoE sliding_window handling and PretrainedConfig pydantic issues are unrelated." }, { "left": "issue:46869", "right": "issue:47228", "accept": false, - "reason": "YOLO multi-GPU mapping and Sam3 compilation regression do not share a code path." + "reason": "YOLO multi-GPU device_map failure and Sam3 bf16 compilation regression are different problems." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "int4 OOM from CUDA memory reservation is unrelated to tied-weights translation model issues." + "reason": "int4 quantization OOM from reserved memory and tied-weights translation-model issues are not the same bug." }, { "left": "issue:45507", "right": "issue:47228", "accept": false, - "reason": "GraniteMoEHybrid invalid method and Sam3 bf16 compilation are different bugs." + "reason": "GraniteMoEHybrid invalid method and Sam3 bf16 compilation regression do not match." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Remote-code breakage and torch 2.4.1/transformers 5.8.0 incompatibility are too broad and not the same defect." + "reason": "Broad remote_code breakage and a specific torch 2.4.1/Transformers 5.8.0 incompatibility are related only at a very high level, not the same concrete issue." }, { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Both are multimodal, but image-preprocessor/model dtype mismatch is not the same as video token/template mismatch." + "reason": "Image-preprocessor dtype mismatch and Qwen3.5 video token/feature mismatch are different multimodal failures." }, { - "left": "issue:41093", - "right": "issue:43596", + "left": "issue:44479", + "right": "issue:46294", "accept": false, - "reason": "Mask length mismatch and ZeRO-3/BertModel index OOB are distinct IndexError scenarios." + "reason": "Both are Qwen video-input regressions, but the affected models and concrete failure modes differ." }, { - "left": "issue:43054", - "right": "issue:43994", + "left": "issue:41093", + "right": "issue:43596", "accept": false, - "reason": "Both involve SigLIP2 quality problems, but one is degraded text embeddings and the other is nonsensical pipeline output; not enough evidence of one bug." + "reason": "Mask-shape mismatch and ZeRO-3 zero-size index error are distinct IndexError classes." }, { "left": "issue:39921", "right": "issue:46869", "accept": false, - "reason": "Gemma3N special-token projection error and YOLO multi-GPU failure are unrelated." + "reason": "Gemma3N special-token projection error and YOLO multi-GPU device mapping are unrelated." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Siglip2 quality regression and Siglip2 nonsensical AutoModel/pipeline output may be related in theme, but the concrete bug is not clearly the same." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Composite-config unwrapping and uninitialized `inv_freq` after `to_empty(cuda)` are separate initialization bugs." + "reason": "Composite Qwen config unwrapping and uninitialized rotary embedding memory are different code paths." }, { "left": "issue:46640", "right": "issue:46661", "accept": false, - "reason": "Flash Attention 2 issue and remote-code `config_class` error do not describe the same change or bug." + "reason": "Flash Attention 2 fallback issue and remote-code config_class AttributeError are unrelated." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "ZeRO-3/BertModel index OOB is unrelated to resume-training with `trust_remote_code`." + "reason": "ZeRO-3 BertModel index error and resume-training with local remote-code model are different failures." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 image dtype mismatch and Qwen3.5 embedding unpacking in TRL are separate bugs." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Custom-model initialization in Jupyter and flash-attn2 fallback gating are different regressions." + "reason": "Jupyter custom-model initialization and flash-attn2 fallback gating are unrelated." }, { "left": "issue:39921", "right": "issue:45507", "accept": false, - "reason": "Gemma3N special-token projection and GraniteMoEHybrid invalid-method errors are unrelated." + "reason": "Gemma3N special-token projection and GraniteMoEHybrid invalid method are not the same change or bug." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "High perplexity on gpt-oss-20b and GLM5 are different model-specific reports." - }, - { - "left": "issue:42915", - "right": "issue:45399", - "accept": false, - "reason": "Qwen3Moe FP8 failure is unrelated to flash-attn2 fallback selection logic." - }, - { - "left": "issue:43404", - "right": "issue:45005", - "accept": false, - "reason": "Both mention tied weights, but one is a Mistral3 lm_head tie bug and the other is a translation-model v5 issue; not the same concrete fix." + "reason": "gpt-oss perplexity anomaly and GLM5 are unrelated issues." } ] }, @@ -1264,6 +1309,7 @@ "issue:41762", "issue:42915", "issue:43054", + "issue:43404", "issue:43493", "issue:43644", "issue:43828", @@ -1284,8 +1330,8 @@ "issue:44928", "issue:45005", "issue:45183", + "issue:45399", "issue:45632", - "issue:45663", "issue:45698", "issue:45701", "issue:45859", @@ -1304,6 +1350,8 @@ "issue:47333" ], "soft_pairs": [ + "issue:42915|issue:45399", + "issue:43404|issue:45005", "issue:45632|issue:45698", "issue:39186|issue:44918", "issue:43873|issue:45005", @@ -1314,11 +1362,11 @@ "issue:44530|issue:45005", "issue:43975|issue:44292", "issue:45183|issue:46463", - "issue:44863|issue:45005", "issue:35308|issue:42915", + "issue:44863|issue:45005", "issue:39692|issue:43054", - "issue:46489|issue:46491", "issue:43950|issue:44534", + "issue:46489|issue:46491", "issue:36331|issue:41762", "issue:43644|issue:43950", "issue:46032|issue:47246", @@ -1332,226 +1380,231 @@ "issue:45987|issue:46858", "issue:43950|issue:45902", "issue:44910|issue:45859", - "issue:43054|issue:43493", - "issue:45663|issue:45902" + "issue:43054|issue:43493" ], - "prepared_review_unit_hash": "cd0eb38c66134c3c8379d57d482353e134309d83486b305cfd56664a1a85c041", + "prepared_review_unit_hash": "2e64fbcd869b23d9278f526ea24b2f23f04e237c1f608af415181dd81be96ea2", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36428, - "estimated_input_tokens": 9107, - "estimated_eval_tokens": 18470 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37309, + "estimated_input_tokens": 9328, + "estimated_eval_tokens": 18912 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is still mostly a semantic false-positive mix of unrelated issue families, but it contains a real duplicate subgroup around the Transformers v5 non-persistent-buffer regression. In that subgroup, #43644 is the earliest report, while #43950 is the clearest representative and #44534 is a follow-up/duplicate.", - "confidence": 0.86, - "canonical_issue_reason": "There is no single cluster-wide canonical issue because the packet splits into multiple unrelated bug families. For the non-persistent-buffer regression subgroup, #43644 is the earliest report in the chain; #43950 is the best-formulated duplicate representative, and #44534 is another close follow-up.", + "summary": "Mostly heterogeneous set, with a few true duplicate subclusters: non-persistent-buffer corruption in v5, input-validation error messaging, cache-length contract violations, Gemma4 quantized embedder dtype casting, Mamba2-family recurrence bugs, and disk-offloaded Qwen3 save failures. The rest are related only at a broad subsystem level and should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "43950 is the best anchor: it states the broad v5 non-persistent-buffer corruption regression and best subsumes the closer follow-ups in this batch.", "canonical_pr_reason": null, - "best_issue_reason": "#43950 is the best overall issue anchor for the non-persistent-buffer regression subgroup because it states the regression most cleanly and generalizes the behavior described by #43644 and #44534.", + "best_issue_reason": "43950 is the most representative issue overall because it has the clearest root-cause framing and the broadest duplicate coverage among the accepted pairs.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Different bugs: one is a FineGrainedFP8Config failure, the other is a flash-attn2 fallback gating issue." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Related to tied weights, but the model families and concrete failure modes differ too much to treat as the same bug." + }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Both are custom-code/model-loading complaints, but cache-path collision and wrong custom-module selection are different bugs." + "reason": "Both involve custom-code loading, but one is a leaf-directory cache collision and the other is wrong module selection after save_pretrained." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "FSDP weight-shape failure and Qwen3.5 input-embedding unpacking failure are different code paths and symptoms." + "reason": "FSDP 2-D weight shape and TRL embedding unpacking are different code paths and symptoms." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior and tied-weight translation-model regressions are not the same defect." + "reason": "Quantization/offloading behavior is not the same as tied-weight regression." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A missing ReLU in an indexer and stale docs for a removed command are unrelated." + "reason": "A DSA indexer discussion and a removed docs command are unrelated." }, { "left": "issue:35308", "right": "issue:36331", "accept": false, - "reason": "IterableDataset multi-GPU crash and compute_loss signature incompatibility are separate trainer issues." + "reason": "IterableDataset multi-GPU crash is unrelated to the custom compute_loss signature break." }, { "left": "issue:35308", "right": "issue:39921", "accept": false, - "reason": "Distributed dataloader crash and Gemma3N special-token projection error are unrelated." + "reason": "No shared underlying failure; training crash vs Gemma3N projection/tokenizer error." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "A cache-strategy typo and docs mentioning a removed command are not the same bug." + "reason": "Just a typo vs a docs/regression issue." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache group-type crash on Qwen3.5 is unrelated to tied-weight regressions." + "reason": "PagedAttentionCache/linear_attention crash is unrelated to tied weights." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Tokenizer detokenization regression and Qwen-3-8B-NVFP4 runtime failure are different problems." + "reason": "Tokenizer detokenization regression and NVFP4 runtime error are different problems." }, { "left": "issue:45183", "right": "issue:46463", - "accept": false, - "reason": "Both involve validation messaging, but the specific input-validation bugs differ and are not clearly the same code-path issue." - }, - { - "left": "issue:44863", - "right": "issue:45005", - "accept": false, - "reason": "NemotronH checkpoint loading failure is unrelated to tied-weight translation-model issues." + "accept": true, + "reason": "46463 looks like the T5-specific manifestation of the broader XOR validation/error-message bug described in 45183." }, { "left": "issue:35308", "right": "issue:42915", "accept": false, - "reason": "IterableDataset distributed-training crash and Qwen3Moe FP8 failure are unrelated." + "reason": "Unrelated training-distribution crash vs FP8 configuration failure." }, { - "left": "issue:39692", - "right": "issue:43054", + "left": "issue:44863", + "right": "issue:45005", "accept": false, - "reason": "Docs-example mismatches/quantization failure and SigLIP2 text-embedding degradation are not the same bug." + "reason": "Model loading failure is not the same as tied-weight translation-model regression." }, { - "left": "issue:46489", - "right": "issue:46491", + "left": "issue:39692", + "right": "issue:43054", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and likely different tokenizer rules." + "reason": "Docs example errors are unrelated to SigLIP2 embedding quality/discrepancy reports." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "These describe the same non-persistent-buffer regression: buffers marked persistent=False are filled/corrupted during load." + "reason": "Same v5 non-persistent-buffer corruption regression; 44534 is a near-duplicate symptom report." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but on different models and different failure modes." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Unexpected compute_loss kwargs and ZeRO-3 loading IndexError are unrelated trainer/model-loading failures." + "reason": "Trainer API signature break vs DeepSpeed ZeRO-3 loading failure are unrelated." }, { "left": "issue:43644", "right": "issue:43950", "accept": true, - "reason": "Same underlying regression in non-persistent buffer handling; #43950 is the more complete formulation of the bug." + "reason": "Same non-persistent-buffer junk/corruption regression in transformers 5.x." }, { "left": "issue:46032", "right": "issue:47246", - "accept": false, - "reason": "Both concern Mamba2-style recurrence, but the concrete code paths differ enough that this is not clearly the same bug." + "accept": true, + "reason": "Both describe the same Mamba2-family recurrence/use_cache bug; 47246 is the NemotronH port that missed the earlier fix." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "bitsandbytes constructor incompatibility and missing Float8 storage lookup are distinct compatibility failures." + "reason": "Different quantization backend incompatibilities with different exception types." }, { "left": "issue:45920", "right": "issue:46489", "accept": false, - "reason": "AutoTokenizer wrong IDs across several models and DeepSeek-Coder v1 detokenization are related only at a broad tokenizer-regression level." + "reason": "Tokenizer-output regressions, but they affect different tokenizers/models and are not the same bug." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "SigLIP2 embedding quality and tokenization-version changes are unrelated." + "reason": "SigLIP2 embedding quality and version-dependent tokenization are unrelated." }, { "left": "issue:46899", "right": "issue:46932", "accept": true, - "reason": "Both report the same Gemma4 quantized vision/audio embedder dtype-casting problem." + "reason": "46932 explicitly says it is the same issue; both point to the unsafe dtype cast in Gemma4 quantized multimodal embedders." }, { "left": "issue:47332", "right": "issue:47333", "accept": true, - "reason": "These are the same offloading-save failure for disk-offloaded Qwen3 models, just on different variants." + "reason": "Same disk-offloaded Qwen3 save failure, just on different Qwen3 model variants." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Unexpected zero probabilities and worse text embeddings are different SigLIP2 symptom reports, not the same concrete failure." + "reason": "Different SigLIP2 symptoms: zero probabilities vs worse text embeddings." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Dtype mismatch under autocast and RLHF NaN collapse from 3D position_ids/SDPA fallback are separate issues." + "reason": "Autocast dtype mismatch is not the same as the Qwen3.5 SDPA/NaN RLHF issue." }, { "left": "issue:45987", "right": "issue:46858", "accept": true, - "reason": "Same get_seq_length contract problem: the cache length is returned as a tensor instead of a plain int, breaking downstream arithmetic." + "reason": "Same get_seq_length contract bug; 46858 is the backend-specific torch.compile manifestation." }, { "left": "issue:43950", "right": "issue:45902", - "accept": false, - "reason": "Shared buffer/materialization theme, but one is a general non-persistent-buffer regression and the other is a specific inv_freq uninitialized-memory bug." + "accept": true, + "reason": "Same v5 non-persistent-buffer/materialization regression; 45902 is a specific buffer-level manifestation." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Flash-Attention illegal-memory access from 3D position_ids and CPU-offload incompatibility in a rotary embedding are different failure modes." + "reason": "Flash-attention packed-sequence crash and CPU-offload rotary-embedding incompatibility are different defects." }, { "left": "issue:43054", "right": "issue:43493", "accept": false, - "reason": "A model-quality complaint and an HF-vs-JAX implementation discrepancy are related only at a very high level." - }, - { - "left": "issue:45663", - "right": "issue:45902", - "accept": false, - "reason": "Gemma4 FSDP2 shared-kv-state rebuild and Qwen3_5Moe inv_freq uninitialized-memory handling are unrelated." + "reason": "Both concern SigLIP2, but one is an embedding-quality complaint and the other is an HF-vs-JAX implementation discrepancy." } ] }, "evaluator_result": { - "accept": false, - "feedback": "Rejecting because the canonical issue rationale is not fully grounded in the packet: #43644 was created before #43950, so #43950 is not the earliest report in that duplicate chain. The rest of the clustering summary is broadly conservative, but the chronology claim is overstated and needs correction." + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The accepted pairs are plausible single-bug subclusters, especially the non-persistent-buffer regression around #43950, the cache-length contract issues, the Gemma4 dtype-cast pair, the Mamba2/NemotronH recurrence bug, and the Qwen3 disk-offload save failures. The rejected pairs are generally framed narrowly enough to avoid overclaiming. No soft-edge verdict appears to overstate the evidence materially." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1571,8 +1624,6 @@ "nodes": [ "issue:24643", "issue:28530", - "issue:36331", - "issue:41093", "issue:41553", "issue:41762", "issue:42491", @@ -1609,16 +1660,19 @@ "issue:45356", "issue:45399", "issue:45488", + "issue:45663", "issue:45685", "issue:45715", "issue:45759", "issue:45823", + "issue:45902", "issue:46315", "issue:46489", "issue:46661", "issue:46869" ], "soft_pairs": [ + "issue:45663|issue:45902", "issue:42947|issue:44387", "issue:44403|issue:44451", "issue:24643|issue:28530", @@ -1628,17 +1682,17 @@ "issue:43643|issue:45020", "issue:44991|issue:45399", "issue:42947|issue:43856", - "issue:43824|issue:43872", "issue:45685|issue:45823", + "issue:43824|issue:43872", "issue:45081|issue:45399", "issue:45020|issue:45823", "issue:43782|issue:44960", "issue:44661|issue:45020", - "issue:41553|issue:43883", "issue:43828|issue:45823", + "issue:41553|issue:43883", "issue:45020|issue:45356", - "issue:42491|issue:44368", "issue:43872|issue:46869", + "issue:42491|issue:44368", "issue:45759|issue:46661", "issue:41553|issue:43881", "issue:45020|issue:46661", @@ -1648,230 +1702,232 @@ "issue:43827|issue:44509", "issue:43881|issue:43883", "issue:43257|issue:45305", - "issue:43792|issue:44220", - "issue:36331|issue:41093" + "issue:43792|issue:44220" ], - "prepared_review_unit_hash": "1413d1388a43cb2626afd80b3ccfaf83c97d8f82f1c6410a2793ab2965865b93", + "prepared_review_unit_hash": "926cef98dec0e51e799188fd6ee0df424eeda13aad5032f9df0346af730661ca", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37376, - "estimated_input_tokens": 9344, - "estimated_eval_tokens": 18944 + "serialized_chars": 37412, + "estimated_input_tokens": 9353, + "estimated_eval_tokens": 18962 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous: most items are adjacent Transformers regressions around model/tokenizer loading, MoE/DeepSpeed, or docs, but they do not look like true duplicates. No soft pairs are safe to merge.", - "confidence": 0.36, - "canonical_issue_reason": "issue:45020 is the broadest, most central regression report in the set, and several soft-similar issues orbit remote_code / model-loading breakages around it.", + "summary": "This cluster is mostly a loose collection of Transformers v5-era regressions around model/tokenizer loading, with only a couple of clear duplicate-like pairs. The strongest canonical issue is the broad remote-code loading regression; the only soft edges I\u2019d accept are the docs pipeline-removal pair and the Whisper/fbank feature-extraction pair.", + "confidence": 0.52, + "canonical_issue_reason": "issue:45020 is the broadest user-facing regression here, centered on remote-code model loading and general enough to sit above the more specific loading/tokenizer failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the strongest single issue to anchor triage: clear user impact, broad scope, and it matches multiple nearby reports without being too narrow.", + "best_issue_reason": "It has the widest scope and best fits the cluster\u2019s dominant theme of v5 loading regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Different bugs: Gemma4 FSDP/shared_kv_states error vs Qwen3_5 MoE inv_freq materialization bug." + }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "Both mention training/memory symptoms, but one is LoRA gradient-checkpointing behavior and the other is int4 quantization OOM; different root causes." + "reason": "Both are memory/efficiency complaints, but they hit different mechanisms: LoRA checkpointing vs int4 reserved memory/OOM." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "One is about reducing noisy loading logs, the other is a hard load failure for a specific model; not the same bug." + "reason": "Generic loading noise vs a specific model load failure; no shared concrete code path." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "Training with DeepSpeed vs early-stopping metric handling are unrelated failures." + "reason": "Unrelated training/deepspeed runtime error vs early-stopping metric-selection issue." }, { "left": "issue:43975", "right": "issue:45237", "accept": false, - "reason": "Tokenizer detokenization corruption and AMD GPU model execution are different code paths." + "reason": "Tokenizer detokenization bug vs AMD GPU model-support failure; different subsystems." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Both involve loading large models with special configs, but the concrete failures (ZeRO-3 empty tensor indexing vs FineGrainedFP8Config) are distinct." + "reason": "Different DeepSpeed/Qwen problems: ZeRO-3 load crash vs FineGrainedFP8Config failure." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Same broad loading area, but different models and different problem statements; one is log noise, the other is an actual load failure." + "reason": "Both are generic loading failures, but they are model-specific and not the same bug." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "Both touch remote_code/config loading, but one is missing fields in AutoConfig return values while the other is a broader regression across recent versions." + "reason": "Remote-code is the common theme, but one is missing fields from AutoConfig and the other is a broad loading regression umbrella." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading failure and flash-attn fallback gating are unrelated." + "reason": "Tokenizer-loading failure for a specific model vs flash-attn fallback gating; unrelated." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Both are memory-related, but one is gradient-checkpointing behavior with PEFT and the other is MoE training memory use; not the same bug." + "reason": "Different symptoms and code paths: gradient checkpointing vs MoE training memory usage." }, { - "left": "issue:43824", - "right": "issue:43872", + "left": "issue:45685", + "right": "issue:45823", "accept": false, - "reason": "Importing a missing class and a bitsandbytes Int8Params incompatibility are separate failures." + "reason": "MPS int histogram limitation is unrelated to Gemma4 device-mapping mismatch." }, { - "left": "issue:45685", - "right": "issue:45823", + "left": "issue:43824", + "right": "issue:43872", "accept": false, - "reason": "Different devices and different errors: MPS histogram support vs device-map mismatch in Gemma4." + "reason": "Import error for a model class vs bitsandbytes Int8Params incompatibility; not the same bug." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer regex patch crash and flash-attn2 fallback blocking are unrelated." + "reason": "Tokenizer regex crash vs flash-attn fallback blocking; different failure points." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "Remote-code loading regressions and Gemma4 device placement are different problems." + "reason": "Broad remote-code loading regressions vs Gemma4 PLE device mismatch; different concrete problems." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight_only loading and GLM5 are different model-specific issues." + "reason": "Qwen3VL weight-only loading error vs a generic GLM5 issue title; insufficient evidence of the same bug." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping registration failure is not the same as the broader remote_code loading regression." - }, - { - "left": "issue:41553", - "right": "issue:43883", - "accept": false, - "reason": "Bad tokenizer error message for Voxtral and missing all_tied_weights_keys on Molmo are unrelated." + "reason": "Tokenizer mapping-name edge case vs broad remote-code regressions; not the same concrete fix." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Autocast dtype mismatch in Phi-tiny-MoE and Gemma4 device mismatch are different runtime issues." + "reason": "Autocast dtype mismatch vs Gemma4 PLE device mismatch; different code paths." }, { - "left": "issue:45020", - "right": "issue:45356", + "left": "issue:41553", + "right": "issue:43883", "accept": false, - "reason": "Both are tokenizer/remote-code-adjacent regressions, but the concrete bug and affected model family differ." + "reason": "Bad error message for AutoTokenizer loading Voxtral vs missing all_tied_weights_keys attribute; unrelated." }, { - "left": "issue:42491", - "right": "issue:44368", + "left": "issue:45020", + "right": "issue:45356", "accept": false, - "reason": "LoRA compatibility across hf4.x/hf5.x and a tie_word_embeddings warning are not the same failure." + "reason": "Remote-code loading regression umbrella vs Kimi tokenizer codec-handling regression; different tokenizer bug." }, { "left": "issue:43872", "right": "issue:46869", "accept": false, - "reason": "bitsandbytes initialization incompatibility and multi-GPU device-map failure are different issues." + "reason": "bitsandbytes constructor incompatibility vs YOLO multi-GPU device-map failure; unrelated." }, { - "left": "issue:45759", + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA compatibility break across versions vs a warning about tie_word_embeddings; not the same underlying bug." + }, + { + "left": "issue:45759", "right": "issue:46661", "accept": false, - "reason": "Composite Qwen config unwrap logic and remote-code model_class/config_class lookup are distinct code paths." + "reason": "Composite multimodal config unwrapping bug vs remote-code model_class config_class AttributeError; different failure modes." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Tokenizer error messaging and GLM-4V loading failure are different bugs." + "reason": "Different loading issues for different models; only superficial similarity in being model-load failures." }, { "left": "issue:45020", "right": "issue:46661", "accept": false, - "reason": "Both involve remote-code models, but one is broad regression coverage and the other is a specific AutoModel attribute error." + "reason": "The latter is a specific remote-code AttributeError, but not clearly the same concrete bug as the broad regression umbrella." }, { "left": "issue:45488", "right": "issue:46489", "accept": false, - "reason": "Both are DeepSeek tokenizer regressions, but they describe different model families and different failure manifestations; too risky to merge." + "reason": "Both involve tokenizer regressions, but they affect different DeepSeek families and appear to need separate fixes." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Tied lm_head metadata and non-persistent buffer corruption are unrelated state-loading bugs." + "reason": "lm_head tying issue vs non-persistent buffer corruption; distinct model-loading bugs." }, { "left": "issue:45715", "right": "issue:46315", "accept": false, - "reason": "One is a per-id loop inefficiency in converting tokens, the other is a slower added-token lookup path; similar theme, different hotspot." + "reason": "One is skip_special_tokens per-id loop inefficiency; the other is convert_tokens_to_ids slowdown with many added tokens." }, { "left": "issue:43827", "right": "issue:44509", - "accept": false, - "reason": "These look like the same docs-family regression around removed pipeline tasks, but not enough evidence to treat as the same exact duplicate." + "accept": true, + "reason": "Both are the same docs cleanup for removed v5 pipeline tasks; title wording differs but the underlying change is identical." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Different models and different load-time failures." + "reason": "Different model-specific loading failures with no shared code-path evidence." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Both mention Qwen/MoE training, but one is weight conversion on load and the other is gradient averaging with DeepSpeed." + "reason": "Qwen3 MoE weight conversion on accelerate+deepspeed vs gradient averaging with GAS; not the same bug." }, { "left": "issue:43792", "right": "issue:44220", - "accept": false, - "reason": "Whisper failing to run and a specific fbank feature-extraction issue may be related, but the evidence is too thin to confirm the same underlying bug." - }, - { - "left": "issue:36331", - "right": "issue:41093", - "accept": false, - "reason": "Custom compute_loss signature mismatch and tensor mask shape mismatch are unrelated errors." + "accept": true, + "reason": "Both point to the same audio feature-extraction path; the Whisper failure and the _torch_extract_fbank_features() issue look like the same underlying bug." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative: it identifies the cluster as a loose mix of v5-era loading/tokenizer regressions, picks a plausible broad canonical issue (issue:45020), and only accepts soft pairs where the titles strongly support a shared fix (docs pipeline removals and the Whisper/fbank audio path). I don\u2019t see an overstatement or an unsafe duplicate claim here." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -1891,6 +1947,7 @@ "issue:24643", "issue:30064", "issue:36010", + "issue:36331", "issue:41093", "issue:41553", "issue:42617", @@ -1938,25 +1995,26 @@ ], "soft_pairs": [ "issue:46894|issue:46897", + "issue:36331|issue:41093", "issue:42947|issue:45923", "issue:43940|issue:45313", - "issue:44479|issue:44560", "issue:24643|issue:30064", + "issue:44479|issue:44560", "issue:43854|issue:44863", "issue:41093|issue:44805", "issue:43366|issue:43828", "issue:43716|issue:46468", "issue:44928|issue:45588", - "issue:44530|issue:44863", "issue:43366|issue:43531", + "issue:44530|issue:44863", "issue:43596|issue:46515", "issue:44863|issue:46402", "issue:45325|issue:45902", "issue:43366|issue:44589", "issue:43638|issue:46515", + "issue:43716|issue:44560", "issue:45663|issue:45823", "issue:41553|issue:44991", - "issue:43716|issue:44560", "issue:42915|issue:43957", "issue:44155|issue:45584", "issue:43824|issue:43957", @@ -1968,172 +2026,178 @@ "issue:44683|issue:46515", "issue:43645|issue:43646" ], - "prepared_review_unit_hash": "679d2b6ce598d0606646053feefc5db9028dec672e2291793a3285a92e505ad7", + "prepared_review_unit_hash": "9b6f4d1cad3e6263af6e6530352fba0b95956b9a7e46efb30934ed6282e2af0f", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36183, - "estimated_input_tokens": 9046, - "estimated_eval_tokens": 18348 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37079, + "estimated_input_tokens": 9270, + "estimated_eval_tokens": 18796 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous cluster of unrelated Transformers issues. Keep 44479 as the anchor because it is the broadest, most versioned regression report, but the Qwen video StopIteration case and the double-shift pair should be treated as related-pattern reports rather than definite duplicates.", - "confidence": 0.74, - "canonical_issue_reason": "Issue 44479 is the best anchor: it describes a concrete v5.3.0 regression and spans multiple Qwen video variants, making it the broadest representative bug report in the set.", + "summary": "The supported duplicates are the two training-loss double-shift reports, the exact mask/tensor shape mismatch reports, and the custom-model-init regression pair. The DeepSpeed/Bert ZeRO-3 reports are not accepted here; the packet does not show a sufficiently shared underlying bug to merge them. Most other soft edges look like title-similarity false positives.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43646 is the cleanest canonical representative for the accepted custom-model-initialization regression: it states the broad Transformers 5.0.0 breakage without the notebook-specific framing in 43645.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44479 is the most suitable global issue because it is actionable, version-specific, and covers the widest affected surface among the issue cluster.", + "best_issue_reason": "Among the accepted issue duplicates, 43646 is the most general and concise issue to anchor the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46894", "right": "issue:46897", + "accept": true, + "reason": "Same training-loss double-shift bug pattern across two models; strong duplicate match." + }, + { + "left": "issue:36331", + "right": "issue:41093", "accept": false, - "reason": "Very similar title pattern, but the packet only shows model-specific double-shift reports and a shared reference pattern; that is not enough to conclude the same underlying bug or same fix." + "reason": "Different failures: trainer compute_loss signature regression vs mask/tensor shape mismatch." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Both involve training/checkpointing language, but one is a PEFT LoRA checkpointing problem and the other is a missing model flag; different code paths and symptoms." + "reason": "Related training features, but different bugs: LoRA checkpointing behavior vs missing model flag." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are ZeRO-3 loading failures, but they affect different Qwen families and describe different missing-weight symptoms without evidence of one shared defect." + "reason": "Both involve ZeRO-3 weight loading, but the model families and failure modes differ too much to merge." }, { - "left": "issue:44479", - "right": "issue:44560", + "left": "issue:24643", + "right": "issue:30064", "accept": false, - "reason": "The broad 5.3.0 Qwen video regression and the specific Qwen3-vl-embedding StopIteration report look related, but the packet does not ground them as the same underlying bug or mergeable fix." + "reason": "Unrelated: DeepSpeed training error vs image processor void-segmentation handling." }, { - "left": "issue:24643", - "right": "issue:30064", + "left": "issue:44479", + "right": "issue:44560", "accept": false, - "reason": "Completely different failures: DeepSpeed training weight-shape error versus image processor void-segmentation-map handling." + "reason": "Both are Qwen video regressions, but the concrete symptoms and code paths differ." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model-loading failures with different targets and symptoms; not enough evidence of a shared underlying bug." + "reason": "Different model-loading problems for different models; not the same defect." }, { "left": "issue:41093", "right": "issue:44805", "accept": true, - "reason": "Near-identical mask/tensor shape mismatch IndexErrors; these read like the same indexing bug reported with different reproduction sizes." + "reason": "Near-identical mask-shape mismatch IndexError; same underlying bug and code path." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "GGUF gpt-oss support request and a torch.autocast dtype-mismatch error are unrelated." + "reason": "GGUF architecture support request vs autocast dtype mismatch are unrelated." }, { "left": "issue:43716", "right": "issue:46468", "accept": false, - "reason": "Different models and different concrete failures: image preprocessor dtype mismatch versus a broken official Nemotron example." + "reason": "Different model families and different loading/forward symptoms." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "RLHF gradient explosion from Qwen3.5 position_ids/SDPA fallback is unrelated to a flash-attention AttributeError on sink-less models." + "reason": "RLHF NaNs from 3D position ids are unrelated to the flash-attention s_aux=None crash." }, { - "left": "issue:44530", - "right": "issue:44863", + "left": "issue:43366", + "right": "issue:43531", "accept": false, - "reason": "PagedAttentionCache linear_attention crash and NemotronH checkpoint loading failure are different code paths and bugs." + "reason": "Different Qwen/GGUF feature and sliding-window issues; not the same bug." }, { - "left": "issue:43366", - "right": "issue:43531", + "left": "issue:44530", + "right": "issue:44863", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window behavior are not the same issue." + "reason": "PagedAttentionCache group-type crash and NemotronH checkpoint loading are unrelated." }, { "left": "issue:43596", "right": "issue:46515", "accept": false, - "reason": "Both are compatibility-style reports, but one is a DeepSpeed zero3/BertModel index error and the other is a torch 2.4.1 incompatibility claim; no shared concrete defect." + "reason": "Specific DeepSpeed/Bert Zero-3 index error vs a broad torch incompatibility report." }, { "left": "issue:44863", "right": "issue:46402", "accept": false, - "reason": "Both involve loading model weights, but they target different model families and failures." + "reason": "Different checkpoint-loading failures for different model families." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Different Qwen models and different bugs: rope-index scaling for still images versus uninitialized inv_freq after meta-to_empty materialization." + "reason": "Different Qwen rope/position-id regressions affecting different paths." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "GGUF architecture support and Float8 storage lookup error are unrelated." + "reason": "GGUF support request and Float8 storage lookup failure are unrelated." }, { "left": "issue:43638", "right": "issue:46515", "accept": false, - "reason": "Zero3 BertModel out-of-bounds index error is not the same as a general torch 2.4.1 incompatibility report." + "reason": "A specific Bert/ZeRO-3 index error is not the same as a generic torch-version incompatibility." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different model families and different video/image processing failures." }, { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Same model family, but different concrete failures: per-layer shared_kv_states rebuild in FSDP2 versus device mismatch under device_map=auto." + "reason": "Gemma4 FSDP2 shared_kv_states KeyError and device-map mismatch are different bugs." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Both are loading-related, but Voxtral AutoTokenizer error messaging and est-roberta tokenizer loading failure are distinct issues." - }, - { - "left": "issue:43716", - "right": "issue:44560", - "accept": false, - "reason": "Both concern Qwen vision/video paths, but one is a dtype mismatch in image preprocessing and the other is a video StopIteration error; not enough evidence they are the same bug." + "reason": "Different tokenizer-loading failures for different models/repos." }, { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and meta-device loading regressions are different code paths and symptoms." + "reason": "Qwen3Moe FP8 config failure is unrelated to meta-device loading regressions." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "AudioFlamingo3 track leakage in batched inference is unrelated to Whisper empty-transcription generation failure." + "reason": "AudioFlamingo3 batch leakage and Whisper empty-transcription handling are unrelated." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "ImportError for a missing Qwen2.5-VL class and meta-device loading failures are distinct problems." + "reason": "Specific import error for one class vs a meta-device loading problem." }, { "left": "issue:42617", @@ -2145,43 +2209,43 @@ "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer corruption on from_pretrained and Kosmos2.5 long OCR index errors are different bugs." + "reason": "Non-persistent buffer corruption is a different regression from long-OCR index errors." }, { "left": "issue:44560", "right": "issue:45685", "accept": false, - "reason": "Video StopIteration in Qwen3-vl-embedding and an MPS histogram implementation error in MoE are unrelated." + "reason": "Qwen video StopIteration and MPS histogram Int support are unrelated." }, { "left": "issue:44560", "right": "issue:44918", "accept": false, - "reason": "Video-input regression and TRL unpacking failure for Qwen3.5 embeddings are different issues." + "reason": "Qwen video input failure and TRL SFT unpacking failure affect different code paths." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import error and Qwen3Moe FP8 configuration failure are unrelated." + "reason": "GenerationMixin import failure and Qwen3Moe FP8 config failure are unrelated." }, { "left": "issue:44683", "right": "issue:46515", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 and torch 2.4.1 incompatibility are different compatibility problems." + "reason": "Compiled flex_attention on torch>=2.9 and a generic torch 2.4.1 incompatibility report are not the same bug." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "Both report the same Transformers 5.0.0 custom-model initialization regression; the notebook case is a narrower reproduction of the broader issue." + "reason": "Same custom-model-initialization regression; 43645 is the notebook-specific manifestation of 43646." } ] }, "evaluator_result": { "accept": false, - "feedback": "The anchor issue choice is grounded, and several false-pair rejections are conservative. However, two accepted soft-edge pairs are a bit too assertive given the packet: 44479\u219444560 is supported only by shared Qwen/video/5.3.0 context, while 44560 is a specific Qwen3-vl-embedding StopIteration report and 44479 is a broader multi-model video regression; 46894\u219446897 are clearly similar, but the evidence in the packet is still only title-level similarity and a shared \u2018double-shift\u2019 pattern label, so calling them the same underlying bug / same fix goes beyond what is directly grounded. I\u2019d treat both as likely-related rather than definite duplicate merges." + "feedback": "The output is mostly conservative, but the summary overstates the evidence: it says the deepspeed/Bert zero-3 reports are among the 'clearest duplicates,' yet both relevant verdicts in the list are marked false. That inconsistency makes the summary insufficiently grounded in the packet. The accepted pairs that are actually supported are the two training-loss double-shift reports, the exact mask-shape mismatch reports, and the custom-model-init reports." }, "evaluator_used": true, "retried": true, @@ -2258,45 +2322,45 @@ "issue:42947|issue:47254", "issue:46468|issue:46869", "issue:45812|issue:46489", - "issue:44661|issue:45823", "issue:44560|issue:44805", + "issue:44661|issue:45823", "issue:44877|issue:45030", "issue:38175|issue:43994", "issue:43638|issue:44661", - "issue:43596|issue:44661", "issue:44661|issue:44805", + "issue:43596|issue:44661", "issue:42491|issue:43720", "issue:45672|issue:46899", - "issue:41720|issue:42915", "issue:43632|issue:45834", + "issue:41720|issue:42915", "issue:43317|issue:44756", "issue:42915|issue:44756", "issue:38175|issue:39692", "issue:43644|issue:44534", "issue:42491|issue:45276", - "issue:44451|issue:44488", "issue:44960|issue:44977", + "issue:44451|issue:44488", "issue:45507|issue:46468", - "issue:44863|issue:46846", "issue:45313|issue:45357", + "issue:44863|issue:46846", "issue:46459|issue:46512", "issue:39692|issue:43994", "issue:44488|issue:44991" ], - "prepared_review_unit_hash": "eae2f4947a86b8f5f8bd06546f926e44976158f7c9c4e779f1a3a2d702b85efa", + "prepared_review_unit_hash": "a7acb913d4bb5d06e282bf2c7a707234764bf44e4f44e3f2aa4a53ebb361e71c", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36337, + "serialized_chars": 36340, "estimated_input_tokens": 9085, "estimated_eval_tokens": 18426 }, @@ -2305,207 +2369,207 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous with one clear exact duplicate pair and one likely same-bug SigLIP2 pair; the rest are separate model/config/runtime regressions rather than duplicates.", - "confidence": 0.41, - "canonical_issue_reason": "No single issue cleanly represents the whole cluster. If forced to anchor it, issue #43994 is the broadest user-facing SigLIP2 report and best captures the wrong-output duplicate with #38175.", + "summary": "The cluster is mostly heterogeneous with a few clear duplicate pairs. I moved the canonical issue anchor away from 44488 to the higher-discussion loading report 41553, and I downgraded the SigLIP2 pair to a non-duplicate because the evidence shows similarity but not the same root cause.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 41553 has the strongest discussion signal in the packet and is a cleaner representative of the loading/error-report subset than the more ambiguous model-specific cases.", "canonical_pr_reason": null, - "best_issue_reason": "#43994 is the best issue to keep as the representative SigLIP2 bug report because it is broader than #38175 and reads like the main user-facing symptom; #43644 is a separate exact-duplicate pair about a different regression.", + "best_issue_reason": "Issue 41553 is the best global issue anchor here because it is the most discussion-heavy and broadly framed, even though the overall cluster still contains several unrelated model-specific bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44910", "right": "issue:46858", "accept": false, - "reason": "Both mention Qwen3.5 and attention/cache failures, but the symptoms and code paths differ: flash-attention packed-sequence misparse vs static-cache slice arithmetic on non-CUDA backends." + "reason": "Different bugs: Qwen3.5 flash-attn/3D position_ids vs static-cache slice arithmetic on non-CUDA backends." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different models and failures: Bamba torch.compile/SDPA vs Qwen3.5 flash-attention memory access." + "reason": "Same broad attention/compile area, but different models, triggers, and failure paths." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Both are tokenizer-loading complaints, but they target different models and error modes; not enough to treat as the same bug." + "reason": "Both are loading-related, but they involve different models and no evidence of the same root cause." }, { "left": "issue:45823", "right": "issue:46515", "accept": false, - "reason": "Gemma4 device-mapping forward mismatch is unrelated to a torch 2.4.1 compatibility report." + "reason": "Device-mapping forward mismatch vs a torch-version incompatibility; unrelated symptoms." }, { "left": "issue:42947", "right": "issue:47254", "accept": false, - "reason": "Both involve PEFT and checkpointing, but one is ineffective gradient checkpointing and the other is a ZeRO-3 CheckpointError; different concrete bugs." + "reason": "Both mention PEFT/checkpointing, but one is ineffective checkpointing and the other is a ZeRO-3 checkpoint error." }, { "left": "issue:46468", "right": "issue:46869", "accept": false, - "reason": "Official Nemotron example breakage is unrelated to YOLO multi-GPU device_map auto failure." + "reason": "Different models and contexts: Nemotron example vs YOLO multi-GPU device_map loading." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Different model families and different tokenizer regressions; too broad to merge." + "reason": "Both are tokenizer regressions, but for different model families and likely separate code paths." }, { - "left": "issue:44661", - "right": "issue:45823", + "left": "issue:44560", + "right": "issue:44805", "accept": false, - "reason": "Tokenizer-mapping helper failure is a different code path from Gemma4 device placement during forward." + "reason": "Different failures on different inputs/models; only a vague runtime-error similarity." }, { - "left": "issue:44560", - "right": "issue:44805", + "left": "issue:44661", + "right": "issue:45823", "accept": false, - "reason": "Both are runtime indexing errors, but the models and failure mechanisms are distinct." + "reason": "Tokenizer-mapping helper failure vs Gemma4 device mismatch during forward." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both are config-validation issues, but they affect different models and likely different validation checks." + "reason": "Both are config/load issues, but they affect different model types and are not the same bug." }, { "left": "issue:38175", "right": "issue:43994", - "accept": true, - "reason": "Same model and same underlying wrong-output bug for SigLIP2; one report is just a narrower symptom of the other." + "accept": false, + "reason": "Related SigLIP2 symptoms, but the packet does not establish the same underlying bug; similarity is only at the model/symptom level." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "DeepSpeed ZeRO-3 initialization failure is unrelated to tokenizer-mapping-name handling." + "reason": "Different code paths: ZeRO-3 Bert initialization vs tokenizer-mapping behavior." }, { - "left": "issue:43596", - "right": "issue:44661", + "left": "issue:44661", + "right": "issue:44805", "accept": false, - "reason": "Different bugs: ZeRO-3/BertModel indexing vs add-new-model-like tokenizer mapping handling." + "reason": "Tokenizer-mapping helper bug vs a mask/tensor shape IndexError; not the same issue." }, { - "left": "issue:44661", - "right": "issue:44805", + "left": "issue:43596", + "right": "issue:44661", "accept": false, - "reason": "Tokenizer mapping helper bug and mask-shape IndexError are unrelated." + "reason": "Different failures: Deepspeed Zero-3 Bert init vs tokenizer-mapping support." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Both mention model-loading behavior, but qwen3_moe LoRA compatibility and BitNet packed-weight unpacking are separate problems." + "reason": "Both loading-related, but one is a Qwen3 MoE LoRA incompatibility and the other is BitNet packed-weight handling." }, { "left": "issue:45672", "right": "issue:46899", - "accept": false, - "reason": "Both are Gemma4 multimodal quantization issues, but the reported failures are different: torch.finfo uint8 TypeError vs unsafe dtype cast in embedders." + "accept": true, + "reason": "Same Gemma4 multimodal 4-bit quantization bug: unsafe dtype handling in embedders/audio-vision paths." }, { - "left": "issue:41720", - "right": "issue:42915", + "left": "issue:43632", + "right": "issue:45834", "accept": false, - "reason": "Different Qwen3-family issues with different triggers and symptoms; not the same concrete bug." + "reason": "Transformers v5 flag regression vs a Kosmos2.5 OCR index error; unrelated." }, { - "left": "issue:43632", - "right": "issue:45834", + "left": "issue:41720", + "right": "issue:42915", "accept": false, - "reason": "_is_hf_initialized flag regression is unrelated to Kosmos2.5 long-ocr input indexing." + "reason": "Different Qwen-family failures with different mechanisms: cuda assert on device mapping vs FineGrainedFP8Config." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "device_map/offload loading failure and mmap OOM avoidance are separate memory/loading issues." + "reason": "Both are memory/loading problems, but device_map offload failure and Strix Halo mmap OOM are distinct bugs." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "FineGrainedFP8Config failure is unrelated to Strix Halo mmap OOM." + "reason": "No meaningful overlap beyond general memory/perf symptoms." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Same model family, but one is wrong probabilities at inference while the other is a documentation/example mismatch plus quantization failure." + "reason": "Same model family, but a docs/example mismatch with quantization failure is not the same as the zero-probabilities inference bug." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "These are clearly the same regression: Transformers v5 fills non-persistent buffers with junk." + "reason": "Same exact v5 regression about non-persistent buffers being filled with junk; phrasing differs, bug is the same." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "LoRA/qwen3_moe hf4.x vs hf5.x compatibility is unrelated to Gemma4 resize_token_embeddings behavior." + "reason": "Different topics: Qwen3 MoE LoRA portability vs Gemma4 embedding resize behavior." }, { - "left": "issue:44451", - "right": "issue:44488", + "left": "issue:44960", + "right": "issue:44977", "accept": false, - "reason": "Both are tokenizer-load failures, but for different model repos and with no evidence they share the same root cause." + "reason": "Different models and failure modes; both are generation issues but not the same code-path problem." }, { - "left": "issue:44960", - "right": "issue:44977", + "left": "issue:44451", + "right": "issue:44488", "accept": false, - "reason": "GLM5 and Qwen3.5 flash-attention generation failures are different model-specific bugs." + "reason": "Both are tokenizer-loading complaints, but for different models with no evidence of the same root cause." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "Invalid method call in GraniteMoEHybrid is unrelated to Nemotron example breakage." + "reason": "Different models and unrelated failures." }, { - "left": "issue:44863", - "right": "issue:46846", + "left": "issue:45313", + "right": "issue:45357", "accept": false, - "reason": "Checkpoint loading failure for NemotronH and missing linear_attn TP plan entries are different issues." + "reason": "Weight-loading ZeRO-3 issue vs save_pretrained visual-encoder key regression; not the same bug." }, { - "left": "issue:45313", - "right": "issue:45357", + "left": "issue:44863", + "right": "issue:46846", "accept": false, - "reason": "ZeRO-3 weight loading and save_pretrained visual-encoder key regression are separate Qwen3.5 problems." + "reason": "NemotronH checkpoint loading vs Qwen3.5 tensor-parallel plan errors; unrelated." }, { "left": "issue:46459", "right": "issue:46512", "accept": false, - "reason": "One is dtype being ignored on composite checkpoint load; the other is config.dtype staying stale after from_config. Related theme, but not the same bug." + "reason": "Both involve dtype, but one is ignored dtype during checkpoint loading and the other is stale config.dtype after from_config." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Both are SigLIP2 reports, but one is a broken docs example and quantization failure, while the other is inference producing nonsensical outputs." + "reason": "Same model family, but docs-example mistakes and actual inference nonsense are not clearly the same underlying bug." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Tokenizer loading fails for different models; the titles suggest a shared class of bug but not the same concrete case." + "reason": "Both are tokenizer-load regressions on v5, but the affected models and failure details differ too much to merge confidently." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is conservative and grounded in the packet. The two accepted soft edges are defensible from the titles/body context, and the rest of the pairwise rejections appropriately avoid overclaiming duplicate status. No PR-mergeability issue applies here." + "accept": false, + "feedback": "The cluster summary is mostly grounded, but two parts overstate evidence. First, the canonical-issue rationale for issue:44488 is not well supported: it does not have the strongest discussion signal in the packet, and its inbound references are not uniquely high compared with other candidates. Second, the accepted soft edge issue:38175 \u2194 issue:43994 is too strong for the evidence shown: both are SigLIP2-related and symptomatically similar, but the packet does not clearly establish the same underlying bug, so that should be treated more conservatively. The other accepted duplicates look more defensible." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2528,7 +2592,6 @@ "issue:42491", "issue:42947", "issue:43257", - "issue:43335", "issue:43404", "issue:43425", "issue:43454", @@ -2547,9 +2610,11 @@ "issue:44291", "issue:44387", "issue:44402", + "issue:44479", "issue:44488", "issue:44756", "issue:44863", + "issue:44877", "issue:44910", "issue:44960", "issue:45276", @@ -2578,8 +2643,8 @@ "issue:42491|issue:45335", "issue:42491|issue:45292", "issue:44164|issue:44291", - "issue:43854|issue:43931", "issue:43856|issue:44387", + "issue:43854|issue:43931", "issue:45488|issue:45812", "issue:44960|issue:46050", "issue:43708|issue:44075", @@ -2590,8 +2655,8 @@ "issue:41720|issue:44155", "issue:45663|issue:45759", "issue:39692|issue:44960", - "issue:44402|issue:44488", "issue:44910|issue:45588", + "issue:44402|issue:44488", "issue:46897|issue:46901", "issue:43873|issue:44387", "issue:43404|issue:43454", @@ -2601,223 +2666,226 @@ "issue:43957|issue:44756", "issue:44079|issue:46829", "issue:44863|issue:46468", - "issue:43335|issue:45663" + "issue:44479|issue:44877" ], - "prepared_review_unit_hash": "bd44e451c3f922adc09993a1ac79b03b1cc4863df530631ed46dc41df922ada8", + "prepared_review_unit_hash": "1829bac9ac2c6ebfae6a801f90bd751af13d9f650bd545555bf3016f6829ad46", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36303, - "estimated_input_tokens": 9076, - "estimated_eval_tokens": 18408 + "serialized_chars": 36843, + "estimated_input_tokens": 9211, + "estimated_eval_tokens": 18678 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: it mixes unrelated model-loading, tokenizer, quantization/offload, training-loss, and forward-pass bugs. I would not collapse any of the soft pairs as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": "No clear canonical issue: the items span many separate bugs across different models and code paths, so there is no single duplicate anchor for the cluster.", + "summary": "This cluster is mostly heterogeneous: most items are distinct model-loading, quantization, or training bugs. The only clear duplicate-style match is the Gemma4/t5gemma `resize_token_embeddings` propagation issue; everything else is too model- or code-path-specific to merge.", + "confidence": 0.73, + "canonical_issue_reason": "issue:45276 is the strongest canonical issue among the true duplicates: it is the more detailed report of the `resize_token_embeddings` propagation bug, and issue:45335 reads like a narrower variant of the same underlying problem.", "canonical_pr_reason": null, - "best_issue_reason": "If a representative is needed, issue 43873 is the broadest and most active user-facing report, but it only covers one quantization/offloading problem and does not subsume the rest of the cluster.", + "best_issue_reason": "issue:43873 is the best overall issue candidate for the set: it has the most discussion/inbound references and is a broad, actively relevant quantization/offloading report, while most other items are isolated model-specific bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Different models and failures: Qwen3 MoE weight conversion vs GLM-4.7-Flash loading in tests." + "reason": "Both are Qwen3/GLM loading failures, but one is a conversion/offload problem and the other is a specific unit-test load mismatch; different failure modes and fixes." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Both are Transformers 5.0 regressions, but one is junk non-persistent buffers and the other is notebook custom-model initialization; different bugs." + "reason": "Both mention Transformers 5 regressions, but one is about junk-filled non-persistent buffers and the other about Jupyter/custom-model initialization; not the same bug." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Both mention Qwen3/quantization/offload themes, but one is training memory inefficiency and the other is offloading behavior; not the same concrete bug." + "reason": "Both involve Qwen3 MoE and memory, but one is general training-memory inefficiency and the other is offloading behavior with quantization; too different to dedupe." }, { "left": "issue:42491", "right": "issue:45335", "accept": false, - "reason": "LoRA HF4.x/HF5.x compatibility is unrelated to Gemma/t5gemma resize_token_embeddings propagation." + "reason": "LoRA hf4.x/hf5.x incompatibility for Qwen3 MoE is unrelated to a Gemma/T5Gemma embedding-resize bug." }, { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Different bug classes: LoRA model compatibility vs resize_token_embeddings not updating output embeddings." + "reason": "Version incompatibility for a Qwen3 MoE LoRA model is not the same underlying issue as `resize_token_embeddings` failing to update embeddings." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "Both involve loading/saving, but one is extra_state serialization and the other is init_empty_weights/_is_hf_initialized constructor args." + "reason": "One is broken `extra_state` save/load handling, the other is `init_empty_weights` rejecting an unexpected `_is_hf_initialized` argument; different code paths." }, { - "left": "issue:43854", - "right": "issue:43931", + "left": "issue:43856", + "right": "issue:44387", "accept": false, - "reason": "Different model-loading failures for different models and root causes." + "reason": "Both are memory-related, but one is training inefficiency in Qwen3 MoE and the other is int4 quantization increasing reserved CUDA memory/OOM; not the same fix." }, { - "left": "issue:43856", - "right": "issue:44387", + "left": "issue:43854", + "right": "issue:43931", "accept": false, - "reason": "Both are memory-related, but one is Qwen3 MoE training inefficiency and the other is int4 reserved-memory OOM under Transformers 5.x." + "reason": "Both are Qwen3-VL loading problems, but one is a unit-test/model-loading failure and the other is a weight-shape mismatch for a specific checkpoint." }, { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and have different tokenizer-path symptoms." + "reason": "Both affect Granite-related tokenization, but one is a pre-tokenizer override regression and the other is incorrect token IDs; related subsystem, different bug." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "GLM5 loading and sinq quantized save/load breakage are unrelated issues." + "reason": "GLM5 is too vague to match the specific sinq save/load breakage; no evidence they are the same underlying issue." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Trainer checkpoint resumption math is unrelated to SGD optimizer arguments not being applied." + "reason": "Trainer checkpoint-resume step calculation and optimizer SGD args being ignored are unrelated." }, { "left": "issue:45276", "right": "issue:45335", - "accept": false, - "reason": "Both are resize_token_embeddings propagation bugs, but they hit different architectures and different embedding targets; not enough evidence for one duplicate." + "accept": true, + "reason": "Both report `resize_token_embeddings` failing to propagate to model-specific embeddings; the second is a narrower manifestation of the same bug pattern." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Torch version compatibility and meta-device model loading are different problems." + "reason": "Torch 2.10 compatibility is not the same as meta-device loading failures on Transformers 5." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "init_empty_weights constructor mismatch is unrelated to mmap OOM on Strix Halo." + "reason": "`init_empty_weights` argument handling and disabling mmap on Strix Halo are unrelated issues." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Qwen3.5 chat crashes and Gemma4Processor missing _tokenizer are different API failures." + "reason": "Qwen3.5 chat crashes and Gemma4Processor missing `_tokenizer` are different model/processor bugs." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "CUDA assert in Qwen3 auto-device-map is unrelated to AudioFlamingo3 batch token/embedding leakage." + "reason": "A800 `cudaErrorAssert` with Qwen3 auto device mapping is unrelated to AudioFlamingo3 batched inference token/embedding leakage." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Gemma-4 FSDP2 KeyError and Qwen 3.5/3.6 text_config unwrapping are different model-internal bugs." + "reason": "FSDP2 `shared_kv_states` rebuilding and `AutoModelForCausalLM.from_config` not unwrapping `text_config` are different composite-model issues." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "SigLIP2 documentation/example issues are unrelated to the GLM5 loading report." + "reason": "SigLIP2 documentation/example errors are unrelated to GLM5 loading behavior." }, { - "left": "issue:44402", - "right": "issue:44488", + "left": "issue:44910", + "right": "issue:45588", "accept": false, - "reason": "Both concern model/tokenizer loading, but the titles point to different models and likely different vocab mismatch causes." + "reason": "Both crash in flash-attention contexts, but one is a packed-sequence misinterpretation from 3D `position_ids` and the other is an `s_aux=None` sink-less-model error." }, { - "left": "issue:44910", - "right": "issue:45588", + "left": "issue:44402", + "right": "issue:44488", "accept": false, - "reason": "Both are flash_attention crashes, but one is a packed-sequence/position_ids bug and the other is s_aux=None handling for sink-less models." + "reason": "A tokenizer vocab-size mismatch for one model is not the same as a broader `cjvt/sleng-bert` loading failure." }, { "left": "issue:46897", "right": "issue:46901", "accept": false, - "reason": "They share a similar 'double-shift' symptom, but they are model-specific training-loss reports with no evidence of the same code path or a single mergeable fix." + "reason": "Both describe a double-shift training-loss pattern, but they are different model implementations and there is no evidence of one shared code path or one mergeable fix." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Offloading/quantization behavior and increased reserved memory under int4 quantization are related only at a high level, not the same bug." + "reason": "Both involve quantization and memory, but one is offloading behavior and the other is increased reserved memory under int4 quantization; not the same bug." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both are weight-tying problems, but they affect different multimodal model classes and different implementations." + "reason": "Both mention untied `lm_head` weights, but they are separate model families and likely separate tie-weight implementations." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing with PEFT LoRA is unrelated to the apex RMSNorm memory leak." + "reason": "Gradient checkpointing ineffectiveness with PEFT LoRA is unrelated to an apex RMSNorm memory leak." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Gemma-4 FSDP2 shared_kv_states KeyError and Qwen3.5 MoE CPU offload incompatibility are different failures." + "reason": "Gemma-4 FSDP2 key errors and Qwen3.5-MoE CPU-offload incompatibility are distinct issues." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "AudioFlamingo3 batch leakage is unrelated to Gemma4 PLE device mismatch during forward." + "reason": "AudioFlamingo3 batch leakage and Gemma4 PLE device mismatch are different bugs in different model paths." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Meta-device loading regressions and mmap OOM on Strix Halo are different issues." + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are unrelated." }, { "left": "issue:44079", "right": "issue:46829", "accept": false, - "reason": "ModelOutput key assignment and VideoPrism output field wiring are different output-shaping bugs." + "reason": "`ModelOutput` key assignment and VideoPrism hidden-state field wiring are different output-structure bugs." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "NemotronH checkpoint loading and the official Nemotron example are separate issues." + "reason": "NemotronH checkpoint loading failure and the official Nemotron example being broken are related only by model family, not the same concrete issue." }, { - "left": "issue:43335", - "right": "issue:45663", + "left": "issue:44479", + "right": "issue:44877", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and Gemma-4 FSDP2 KeyError are unrelated model-specific bugs." + "reason": "Video-input regression for Qwen VL models and strict config blocking `granite_speech` loading are separate bugs." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet and stays conservative. The best-issue choice is supported by discussion/inbound counts, and the only accepted soft-edge pair (issue:45276 vs issue:45335) is a plausible same-bug duplicate within the resize_token_embeddings propagation family. The remaining pairs are reasonably rejected as model- or code-path-specific rather than clearly mergeable duplicates." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -2838,6 +2906,7 @@ "issue:39404", "issue:39921", "issue:42796", + "issue:43335", "issue:43381", "issue:43425", "issue:43611", @@ -2858,7 +2927,6 @@ "issue:44683", "issue:44756", "issue:44805", - "issue:44877", "issue:44912", "issue:44928", "issue:45084", @@ -2884,9 +2952,9 @@ "issue:47448" ], "soft_pairs": [ - "issue:44479|issue:44877", - "issue:44387|issue:44756", + "issue:43335|issue:45663", "issue:39921|issue:43425", + "issue:44387|issue:44756", "issue:44292|issue:44912", "issue:36010|issue:39404", "issue:45276|issue:45292", @@ -2894,15 +2962,15 @@ "issue:43957|issue:47332", "issue:43381|issue:44928", "issue:45588|issue:45663", - "issue:42796|issue:46326", "issue:43645|issue:44164", "issue:43425|issue:47228", + "issue:42796|issue:46326", "issue:44589|issue:44756", - "issue:45292|issue:45335", "issue:43425|issue:43828", + "issue:45292|issue:45335", "issue:43854|issue:45216", - "issue:44402|issue:44451", "issue:43646|issue:43708", + "issue:44402|issue:44451", "issue:45684|issue:45698", "issue:43425|issue:46869", "issue:46894|issue:46901", @@ -2916,220 +2984,226 @@ "issue:44589|issue:45084", "issue:45823|issue:47448" ], - "prepared_review_unit_hash": "b0c350b4522e33576cfa262e64116d35cf7c946ffeb3e447b8ce7e98613d19a4", + "prepared_review_unit_hash": "429467ae9ec251cbd2abb8c2471f6468434ae86de09d43cec1be92e655b70753", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36926, - "estimated_input_tokens": 9232, - "estimated_eval_tokens": 18720 + "serialized_chars": 36976, + "estimated_input_tokens": 9244, + "estimated_eval_tokens": 18744 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a similarity cluster around unrelated Transformers regressions. The only clear duplicate set is the resize_token_embeddings/output_embeddings family; most other pairs share broad subsystem wording but appear to be distinct bugs or model-specific regressions.", - "confidence": 0.64, - "canonical_issue_reason": "issue:45292 is the cleanest generic report of the resize_token_embeddings propagation bug, and it aligns with the two closely related variants in the cluster.", + "summary": "Most candidate pairs are false positives across unrelated model-, tokenizer-, or runtime-specific regressions. The only clear duplicate cluster is the resize_token_embeddings bug: issue 45292 is the generic report, while 45276 and 45335 are model-specific manifestations.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45292 most cleanly states the shared underlying regression: `resize_token_embeddings` fails to propagate updates to output embeddings. It is the broadest representative of the accepted duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45292 is the most representative triage target here: broad symptom, least model-specific wording, and the most likely hub for the accepted duplicate set.", + "best_issue_reason": "Issue 45292 is the best issue candidate because it describes the core bug generically, without narrowing it to one architecture, while the other accepted issues are specific instances of the same failure mode.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44479", - "right": "issue:44877", + "left": "issue:43335", + "right": "issue:45663", "accept": false, - "reason": "Different bugs and code paths: Qwen video-input regression vs granite_speech config loading." + "reason": "Different models and different failures: sparse-layer construction vs Gemma4 FSDP attention state rebuilding." }, { - "left": "issue:44387", - "right": "issue:44756", + "left": "issue:39921", + "right": "issue:43425", "accept": false, - "reason": "One is an int4 quantization OOM/regression; the other is mmap-related OOM on Strix Halo." + "reason": "Special-token projection error in Gemma3N is unrelated to a Torch 2.10 compatibility issue." }, { - "left": "issue:39921", - "right": "issue:43425", + "left": "issue:44387", + "right": "issue:44756", "accept": false, - "reason": "Gemma3N special-token projection error vs a Torch 2.10 compatibility issue; not the same bug." + "reason": "Both mention OOM, but the causes differ: int4 quantization memory growth vs disabling mmap on Strix Halo." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Different model/quantization failures: Qwen-3-8B-NVFP4 loading vs git-oss-20b MXFP4 fallback." + "reason": "Different quantization paths and symptoms (`NVFP4` run failure vs `MXFP4` fallback to bf16) on different models." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "ImportError for GenerationMixin vs Whisper pipeline return_language behavior; unrelated symptoms." + "reason": "ImportError for `GenerationMixin` is unrelated to Whisper pipeline `return_language` behavior." }, { "left": "issue:45276", "right": "issue:45292", "accept": true, - "reason": "Same underlying resize_token_embeddings propagation bug affecting output embeddings; 45276 is just the Gemma4-specific variant." + "reason": "Same underlying `resize_token_embeddings` propagation bug; 45276 is a Gemma4-specific manifestation of the generic output-embedding issue in 45292." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "save/from_pretrained extra_state handling vs Strix Halo mmap OOM; different problem domains." + "reason": "`extra_state` serialization failure and mmap-induced OOM are different save/load problems." }, { "left": "issue:43957", "right": "issue:47332", "accept": false, - "reason": "Meta-device model loading breakage vs inability to save a disk-offloaded model; not the same code-path issue." + "reason": "Meta-device loading regressions and disk-offloaded save failures are different code paths." }, { "left": "issue:43381", "right": "issue:44928", "accept": false, - "reason": "Gradient checkpointing in eval mode vs RLHF NaNs from 3D position_ids/SDPA fallback; distinct regressions." + "reason": "Gradient checkpointing in eval mode is unrelated to RLHF NaNs from 3D position ids and SDPA fallback." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "flash_attention.py s_aux=None crash vs Gemma4 FSDP2 shared_kv_states KeyError; different fixes needed." - }, - { - "left": "issue:42796", - "right": "issue:46326", - "accept": false, - "reason": "Tokenizer BPE marker handling vs continue_final_message behavior; similar area, different bug." + "reason": "Flash-attention sink-less `s_aux=None` crash is not the same bug as Gemma4 FSDP2 shared KV state rebuilding." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Custom-model initialization in Jupyter vs save/from_pretrained extra_state handling; not the same issue." + "reason": "Custom-model notebook initialization regression is different from `save/from_pretrained` extra_state handling." }, { "left": "issue:43425", "right": "issue:47228", "accept": false, - "reason": "Torch version incompatibility vs Sam3 bf16 compilation regression; only both version-related at a high level." + "reason": "Both mention torch/version sensitivity, but one is broad incompatibility and the other is a specific bf16 compilation regression." }, { - "left": "issue:45292", - "right": "issue:45335", - "accept": true, - "reason": "Same resize_token_embeddings regression, now surfacing at decoder.embed_tokens in t5gemma." + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer BPE-marker handling and `continue_final_message` behavior are different tokenizer bugs." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage deserialization error is unrelated to the Strix Halo mmap OOM issue." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "Generic Torch incompatibility vs Phi-tiny-MoE autocast dtype mismatch; different concrete failure." + "reason": "Torch 2.10 incompatibility is not the same as Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same `resize_token_embeddings` regression, with 45335 showing the decoder.embed_tokens variant of the output-embedding propagation failure." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "GLM-4.7-Flash unit-test loading failure vs Qwen3.5 save_pretrained regression; unrelated." + "reason": "Different model-loading/save regressions on different models and APIs." }, { - "left": "issue:44402", - "right": "issue:44451", + "left": "issue:43646", + "right": "issue:43708", "accept": false, - "reason": "Tokenizer vocab-size mismatch vs inability to load ScandiBERT; both loading-related but not the same root cause." + "reason": "Custom model initialization regression is unrelated to Trainer `resume_from_checkpoint` step \u0570\u0561\u0577\u057eculation." }, { - "left": "issue:43646", - "right": "issue:43708", + "left": "issue:44402", + "right": "issue:44451", "accept": false, - "reason": "Custom model init breakage vs Trainer max_steps miscalculation on resume_from_checkpoint; different bug classes." + "reason": "The ScandiBERT load failure may be related in theme, but the titles do not establish the same special-token/vocab bug." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Read-only permission propagation in save_pretrained vs loading the wrong custom module after save_pretrained." + "reason": "Read-only permission propagation from `save_pretrained` is different from loading the wrong custom module after save." }, { "left": "issue:43425", "right": "issue:46869", "accept": false, - "reason": "Torch 2.10 incompatibility vs YOLO multi-GPU device_map='auto' failure; only broadly version/device related." + "reason": "Torch version incompatibility and multi-GPU `device_map=\"auto\"` failure are different issues." }, { "left": "issue:46894", "right": "issue:46901", "accept": false, - "reason": "Both mention double-shift loss bugs, but they are model-specific reports with no clear shared code path or single change." + "reason": "Both are training-loss shift bugs, but they are model-specific reports without enough evidence of the same code path." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "PagedAttentionCache linear_attention crash vs compile-time non-template node error; unrelated." + "reason": "PagedAttentionCache group-type failure is unrelated to the compiler error about non-template nodes." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Wrong Granite token IDs vs a performance regression in convert_tokens_to_ids; not the same defect." + "reason": "Wrong token IDs for Granite models and slow `convert_tokens_to_ids` are different tokenizer regressions." }, { "left": "issue:46315", "right": "issue:46491", "accept": false, - "reason": "Token conversion slowdown vs CodeLlama leading-space round-trip regression; different symptoms." + "reason": "Performance regression in token conversion and CodeLlama spacing round-trip bug are unrelated." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention torch>=2.9 breakage vs Gemma4 PLE device mismatch; separate issues." + "reason": "Compiled flex_attention on torch >= 2.9 and Gemma4 device-mapping mismatch are different failures." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language regression vs version-dependent tokenization changes; not duplicates." + "reason": "Whisper pipeline regression is unrelated to a general tokenization change report." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "base_model_prefix loading failure vs custom model initialization in notebooks; similar release scope, different bug." + "reason": "Base-model-prefix loading breakage and Jupyter custom-model initialization breakage are not the same concrete bug." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Qwen video-input regression vs mask/tensor shape IndexError; no shared underlying change." + "reason": "Video-input regression for Qwen VL models is unrelated to the mask/tensor shape mismatch." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Missing Float8 storage type vs compile non-template nodes; unrelated failures." + "reason": "Float8 storage lookup failure is not the same as the compiler rejecting non-template nodes." }, { "left": "issue:45823", "right": "issue:47448", "accept": false, - "reason": "Gemma4 device-map mismatch during forward vs unrecognized Gemma4 12B architecture; likely adjacent but not the same bug." + "reason": "Gemma4 device mismatch during forward is unrelated to the model-architecture recognition failure." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative on the soft edges. The accepted resize_token_embeddings pairings are well supported by the issue titles and look mergeable as one bugfix family. The other rejects do not overstate duplication based on the available evidence." + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate cluster is the resize_token_embeddings/output_embeddings trio (45292 with 45276 and 45335), and the other pair rejections are appropriately cautious given only the issue titles." }, "evaluator_used": true, "retried": false, @@ -3158,6 +3232,7 @@ "issue:42915", "issue:43404", "issue:43452", + "issue:43541", "issue:43596", "issue:43611", "issue:43632", @@ -3206,8 +3281,8 @@ "issue:36032|issue:43452", "issue:39404|issue:46665", "issue:43901|issue:44509", - "issue:46050|issue:46326", "issue:44991|issue:45081", + "issue:46050|issue:46326", "issue:44671|issue:44991", "issue:43906|issue:46326", "issue:43906|issue:44530", @@ -3217,220 +3292,227 @@ "issue:44756|issue:47333", "issue:43596|issue:44683", "issue:43856|issue:44756", - "issue:44488|issue:44987", "issue:43632|issue:47050", - "issue:36032|issue:42915", + "issue:44488|issue:44987", "issue:36032|issue:39921", + "issue:36032|issue:42915", "issue:44488|issue:45081", "issue:43716|issue:45084", "issue:42796|issue:43404", "issue:43746|issue:47331", "issue:43611|issue:43644", + "issue:43541|issue:45084", "issue:44462|issue:46489", "issue:45479|issue:46894" ], - "prepared_review_unit_hash": "8d9f156336a90e66e2473b5095349962b647373ddd0450ea502c08d4a9526823", + "prepared_review_unit_hash": "9b9466877ed34d99778f64017fa125a23da06153b4fa67f42c0ce524617622a3", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 36101, - "estimated_input_tokens": 9026, - "estimated_eval_tokens": 18308 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37001, + "estimated_input_tokens": 9251, + "estimated_eval_tokens": 18758 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated issue reports that only share broad Transformers v5/load/tokenizer wording. I do not see any soft pair that clearly describes the same underlying bug or fix.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43611 is the clearest broad regression in the set: it describes a concrete v5 loading break around `base_model_prefix`, which is the most representative of the cluster\u2019s load/init failures.", + "summary": "This cluster is noisy and mostly heterogeneous: many items are Transformers v5 regressions, but they span different code paths (tokenizer loading, model loading, quantization, docs, generation, export, and model outputs). The only clearly close pairings are still not strong enough to treat as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "issue:43611 is the broadest and most central v5 loading-regression report in the set; it best matches the cluster\u2019s recurring theme of model/tokenizer loading breakage, even though the overall cluster is not a clean duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43611 has the strongest combination of specificity and generality for this set: a concrete, reproducible v5 loading regression with a plausible shared root in model initialization.", + "best_issue_reason": "issue:43611 is the best single representative anchor for triage because it describes a concrete, high-level loading failure in Transformers v5 and is closer to the shared pattern than the more niche reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "Different failure modes: extra_state save/load handling vs `from_config` not unwrapping `text_config`." + "reason": "Different failures: extra_state save/load handling vs. from_config text_config unwrapping for multimodal Qwen configs." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Same broad model/output area, but one is a SigLIP2 inference issue and the other is `ModelOutput` key assignment semantics." + "reason": "Different bug classes: wrong SIGLIP2 outputs vs. ModelOutput key assignment when a key was previously None." }, { "left": "issue:40990", "right": "issue:47331", "accept": false, - "reason": "Unrelated: perplexity on gpt-oss-20b vs Metal quantized batched generation corruption." + "reason": "Unrelated problem surfaces: perplexity evaluation anomaly vs. Metal quantization batched generate corruption." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "Different code paths: `torch.export` with `torch_compilable_check` vs Perceiver interpolation at non-default resolution." + "reason": "Both are export/vision-related, but the concrete code paths differ: torch.export with torch_compilable_check vs. Perceiver interpolation handling." }, { "left": "issue:36032", "right": "issue:43452", "accept": false, - "reason": "Both involve loading, but one is a T5 tokenizer special-tokens conflict and the other is `gguf_file` handling for model/tokenizer loading." + "reason": "Both involve loading, but one is a T5Tokenizer add_special_tokens conflict and the other is gguf_file resolution for AutoTokenizer/AutoModel." }, { "left": "issue:39404", "right": "issue:46665", "accept": false, - "reason": "One is Whisper pipeline `return_language`; the other is generic `from_pretrained` behavior inside Python classes." + "reason": "Whisper pipeline return_language regression is unrelated to inconsistent from_pretrained behavior inside Python classes." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs-related, but they cover different API changes and different removed/renamed pipeline behavior." + "reason": "Both are docs issues, but they concern different removed/deprecated pipeline/task references." }, { - "left": "issue:46050", - "right": "issue:46326", + "left": "issue:44991", + "right": "issue:45081", "accept": false, - "reason": "Different subsystems: quantized save/load round-trip vs `continue_final_message` for mistral_common tokenizers." + "reason": "Both are tokenizer-loading bugs, but one is a repo-specific tokenizer load failure and the other is a Mistral regex patch crash in a distinct helper path." }, { - "left": "issue:44991", - "right": "issue:45081", + "left": "issue:46050", + "right": "issue:46326", "accept": false, - "reason": "Both tokenizer-load regressions, but different models and different bugs (`est-roberta` load failure vs Mistral regex patch crashing)." + "reason": "Quantized save/load breakage is not the same as continue_final_message behavior for mistral_common tokenizers." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT masked-LM predictions in v5 is not the same as tokenizer loading for EMBEDDIA/est-roberta." + "reason": "CamemBERT MLM prediction regression and est-roberta tokenizer loading failure are different model behaviors and different code paths." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "No clear overlap beyond tokenizer-related wording; one is a reproduction of issue #38071, the other is a mistral_common `continue_final_message` bug." + "reason": "A reproduction of an external issue is not the same as a Mistral tokenizer continue_final_message bug." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "Different problems: isolated repro of #38071 vs PagedAttentionCache failing on Qwen3.5 linear_attention groups." + "reason": "Different subsystems: reproduction of a prior bug vs. PagedAttentionCache linear_attention crash on Qwen3.5." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper `return_language` pipeline behavior is unrelated to MXFP4 quantization fallback on gpt-oss-20b." + "reason": "Whisper pipeline language-return behavior is unrelated to gpt-oss MXFP4 quantization fallback." }, { "left": "issue:45823", "right": "issue:46899", "accept": false, - "reason": "Same family name (Gemma4) but different bugs: device-map mismatch in forward vs unsafe dtype cast in multimodal embedders under 4-bit quantization." + "reason": "Both mention multimodal dtype/device issues, but one is Gemma4 PLE device_map forwarding and the other is unsafe dtype casting in quantized embedders." }, { "left": "issue:44164", "right": "issue:46821", "accept": false, - "reason": "Different APIs and failures: extra_state save/load vs `window_size` raising `AttributeError` in TimesFM 2.5." + "reason": "save/from_pretrained extra_state handling is unrelated to a TimesFM window_size AttributeError." }, { "left": "issue:44756", "right": "issue:47333", "accept": false, - "reason": "Unrelated: mmap/OOM on Strix Halo vs saving a disk-offloaded Qwen3 model." + "reason": "Both are resource/offload related at a high level, but one is mmap OOM on Strix Halo and the other is disk-offloaded Qwen save failure." }, { "left": "issue:43596", "right": "issue:44683", "accept": false, - "reason": "Different bugs in different stacks: DeepSpeed ZeRO3/BertModel indexing vs `torch.export`/`flex_attention` compilation." + "reason": "Zero3/BertModel indexing failure is a different issue than compiled flex_attention failing on torch>=2.9." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Training memory inefficiency in Qwen3 MoE is not the same as disabling mmap to avoid OOM on Strix Halo." + "reason": "Qwen3 MoE training memory inefficiency is not the same as mmap-induced OOM on Strix Halo." }, { - "left": "issue:44488", - "right": "issue:44987", + "left": "issue:43632", + "right": "issue:47050", "accept": false, - "reason": "Both are load failures, but for different repositories/models and different failure modes." + "reason": "An _is_hf_initialized flag regression is unrelated to CanineModel fp16 ONNX attention dtype mismatch." }, { - "left": "issue:43632", - "right": "issue:47050", + "left": "issue:44488", + "right": "issue:44987", "accept": false, - "reason": "`_is_hf_initialized` flag breakage is unrelated to an ONNX export attention dtype mismatch." + "reason": "Both are loading failures, but one is a specific repo not loading and the other is a different transformers>=5.1.0 regression for another model." }, { "left": "issue:36032", - "right": "issue:42915", + "right": "issue:39921", "accept": false, - "reason": "Tokenizer special-tokens conflict vs Qwen3Moe FineGrainedFP8Config failure are distinct issues." + "reason": "T5 tokenizer method conflict and Gemma3N special-token projection error are different tokenizer/model mechanisms." }, { "left": "issue:36032", - "right": "issue:39921", + "right": "issue:42915", "accept": false, - "reason": "Both mention tokenizer/special tokens, but the underlying bugs differ: T5 tokenizer method conflict vs Gemma3N projection error when adding tokens." + "reason": "T5 tokenizer loading conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Different models and different load paths; no evidence they share a root cause." + "reason": "Different tokenizer load issues: repo-specific loading failure vs. Mistral regex patch crash." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Image preprocessor/model dtype mismatch is unrelated to a template compilation TypeError." + "reason": "Mistral-3 dtype mismatch in image preprocessing is unrelated to a template compilation TypeError." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Tokenizer BPE marker handling and Mistral3 lm_head weight tying are separate model-specific bugs." + "reason": "Tokenizer BPE-marker handling and lm_head weight tying are different bugs in different model components." }, { "left": "issue:43746", "right": "issue:47331", "accept": false, - "reason": "PEFT local checkpoint loading and Metal quantized batched generation corruption do not share a concrete code path." + "reason": "GraniteSpeech PEFT checkpoint loading and Metal quantized batched generate corruption are unrelated." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Both are v5 regressions, but one is `base_model_prefix` loading and the other is non-persistent buffer initialization; distinct fixes." + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; not the same bug." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Mixtral grouped_mm tracing failure is unrelated to the non-template nodes compile error." }, { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Repository tokenizer.json being ignored is not the same as DeepSeek-Coder v1 tokenizer output being wrong in v5+." + "reason": "AutoTokenizer ignoring tokenizer.json is not the same as a DeepSeek-Coder tokenizer-output regression tied to a prior fix gap." }, { "left": "issue:45479", "right": "issue:46894", "accept": false, - "reason": "Different training-loss bugs in different model families; only a superficial pattern match, not the same underlying defect." + "reason": "Degenerate zero-loss classification bug and training-loss double-shift bug are different model/task errors." } ] }, @@ -3468,13 +3550,12 @@ "issue:43421", "issue:43452", "issue:43482", - "issue:43541", "issue:43582", "issue:43632", "issue:43644", "issue:43645", "issue:43646", - "issue:43746", + "issue:43792", "issue:43873", "issue:43931", "issue:44206", @@ -3498,15 +3579,15 @@ "issue:46304", "issue:46326", "issue:46402", + "issue:46869", "issue:46897", "issue:47013", "issue:47312", "issue:47331" ], "soft_pairs": [ - "issue:46285|issue:46402", - "issue:43541|issue:45084", "issue:44661|issue:44683", + "issue:46285|issue:46402", "issue:43299|issue:43931", "issue:42915|issue:44530", "issue:45479|issue:46897", @@ -3517,233 +3598,236 @@ "issue:44737|issue:45859", "issue:40444|issue:46326", "issue:40990|issue:43421", - "issue:42175|issue:46178", "issue:34567|issue:40444", - "issue:43644|issue:43646", + "issue:42175|issue:46178", "issue:45084|issue:45507", - "issue:44206|issue:44220", + "issue:43644|issue:43646", "issue:43582|issue:43632", "issue:44987|issue:44991", "issue:39404|issue:44671", + "issue:44206|issue:44220", "issue:43452|issue:43482", "issue:43482|issue:47312", "issue:43262|issue:47331", - "issue:43208|issue:47013", - "issue:44811|issue:45584", "issue:43582|issue:47331", - "issue:43746|issue:46326", - "issue:43317|issue:43482", - "issue:42915|issue:45084" + "issue:44811|issue:45584", + "issue:43208|issue:47013", + "issue:42915|issue:45084", + "issue:43792|issue:46869", + "issue:43317|issue:43482" ], - "prepared_review_unit_hash": "79b53e3333b9025133ffbb6308f29ff67830083552884bca3a86aed246b0a412", + "prepared_review_unit_hash": "a66139a311f53acac9a40be05a5978338903526336d9fb12ba7d66651eb552ed", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36710, - "estimated_input_tokens": 9178, - "estimated_eval_tokens": 18612 + "soft_pair_count": 29, + "serialized_chars": 36279, + "estimated_input_tokens": 9070, + "estimated_eval_tokens": 18396 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose set of Transformers v5 regressions and model-specific failures, not one duplicate family. The broadest representative is the custom model initialization regression, while most other items are narrower and unrelated.", - "confidence": 0.88, - "canonical_issue_reason": "issue 43646 is the broadest, most representative v5 regression here: a general custom-model initialization breakage that fits the cluster\u2019s overall theme better than the narrower model-specific reports.", + "summary": "This cluster is mostly heterogeneous: many items are Transformers v5 regressions, but they concern different models, subsystems, and failure modes. Only the GGUF loading reports look like the same underlying bug; the rest should stay separate.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43646 is the broadest representative of the v5 regression bucket, but it is still only one specific custom-model-init failure and not a safe umbrella duplicate for the whole cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue 43646 is the best single issue to keep as the cluster representative because it is generic and central, whereas most other items are isolated failures in specific models, loaders, or backends.", + "best_issue_reason": "issue:43646 is the best representative issue because it is the most generic v5 breakage in the set, though it should not absorb the unrelated tokenizer/audio/quantization/offload reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:46285", - "right": "issue:46402", - "accept": false, - "reason": "Different models and failure modes: CLIP text-model structure vs Sam3 text encoder weight loading." - }, - { - "left": "issue:43541", - "right": "issue:45084", + "left": "issue:44661", + "right": "issue:44683", "accept": false, - "reason": "Both are compile/tracing errors, but they hit different code paths and symptoms; not the same bug." + "reason": "Different code paths: tokenizer mapping/model registration vs compiled flex_attention on torch >= 2.9." }, { - "left": "issue:44661", - "right": "issue:44683", + "left": "issue:46285", + "right": "issue:46402", "accept": false, - "reason": "Tokenizer-mapping generation bug vs flex_attention compile failure on newer torch; unrelated." + "reason": "Both are model-loading regressions, but for different models and different root causes." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Different Qwen model families and different loading errors; no clear shared root cause." + "reason": "Qwen loading failures, but one is FP8 config handling and the other is a weight-shape mismatch for a VL model." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Both mention Qwen3 variants, but one is FP8 config loading and the other is PagedAttention cache execution." + "reason": "Both are Qwen-related, but one is FineGrainedFP8Config loading and the other is a cache/group-type crash in PagedAttentionCache." }, { "left": "issue:45479", "right": "issue:46897", "accept": false, - "reason": "Sequence-classification loss bug vs Florence2 training-loss shift bug; distinct problems." + "reason": "Both affect training loss, but they describe different bugs in different models and should not be merged." }, { "left": "issue:43873", "right": "issue:47331", "accept": false, - "reason": "Quantized offload issues on different backends and different runtime failures; not one concrete bug." + "reason": "Quantization/offload problems, but one is GPU offloading behavior and the other is a Metal batched-generate corruption bug." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading failures, but they concern different models and different breakpoints." + "reason": "Both are loading regressions, but they target different models and don't point to the same loader bug." }, { "left": "issue:44530", "right": "issue:46304", "accept": false, - "reason": "Qwen3.5 cache crash vs distributed build regression; unrelated layers." + "reason": "Unrelated failures: Qwen3.5 cache handling vs a distributed-build/PyTorch support regression." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "Tokenizer method-name conflict vs custom model initialization in notebooks; different bugs." + "reason": "Tokenizer method-name conflict vs custom model initialization in notebooks; not the same bug." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Both are device/offload-related, but one is XLNet positional encoding and the other is Qwen3 Moe rotary embedding." + "reason": "Both touch device/offload behavior, but one is CPU placement in XLNet and the other is Qwen3_5Moe offload compatibility." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Multimodal training with multiple images vs mistral_common chat-template handling; different code paths." + "reason": "Different subsystems and symptoms: multimodal dataset finetuning vs mistral_common chat-template behavior." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity anomaly on a model vs runtime special-token post-processor update; not the same issue." - }, - { - "left": "issue:42175", - "right": "issue:46178", - "accept": false, - "reason": "Packaging dependency omission vs backend registry validation error; separate problems." + "reason": "Perplexity quality issue vs tokenizer post-processor update behavior; not the same underlying bug." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "Trainer token-count tracking vs Qwen2.5-VL iterable dataset handling; unrelated." + "reason": "Trainer metrics tracking vs multimodal IterableDataset handling; unrelated." }, { - "left": "issue:43644", - "right": "issue:43646", + "left": "issue:42175", + "right": "issue:46178", "accept": false, - "reason": "Both are v5 init regressions, but one is junk non-persistent buffers and the other is custom model initialization; not identical." + "reason": "Both mention TensorFlow backends, but one is packaging extras and the other is a missing backend mapping entry." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Different runtime errors in different subsystems; no evidence of a shared underlying bug." + "reason": "Different failures: template compilation error vs an invalid method call in GraniteMoEHybrid." }, { - "left": "issue:44206", - "right": "issue:44220", + "left": "issue:43644", + "right": "issue:43646", "accept": false, - "reason": "Both are audio-feature related, but the titles point to different failing functions and symptoms." + "reason": "Both are v5 regressions, but one is non-persistent buffer initialization and the other is custom model initialization in notebooks." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "Apple Silicon allocator TypeError vs _is_hf_initialized flag breakage; unrelated." + "reason": "AppleSilicon allocator TypeError and _is_hf_initialized flag breakage are separate issues." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are loading regressions after v5, but they affect different models and likely different loader paths." + "reason": "Both are model-loading regressions after v5, but they involve different models and likely different loader paths." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper return_language pipeline regression vs CamemBERT masked-LM prediction bug; different models and behavior." + "reason": "Whisper pipeline regression vs CamemBERT masked-LM prediction bug; unrelated." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Same audio area, but the titles point to different feature-extraction failures without enough evidence of one root cause." }, { "left": "issue:43452", "right": "issue:43482", - "accept": false, - "reason": "Both mention GGUF loading, but one is a broad auto-class gguf_file regression and the other is a specific model-loading failure; not enough to assume the same bug." + "accept": true, + "reason": "Both point to the same GGUF loading regression in transformers v5; the Qwen2.5 case looks like a concrete instance of the broader gguf_file breakage." }, { "left": "issue:43482", "right": "issue:47312", "accept": false, - "reason": "GGUF loading failure vs kernel compatibility issue; unrelated." + "reason": "GGUF loading failure and kernel version incompatibility are unrelated." }, { "left": "issue:43262", "right": "issue:47331", "accept": false, - "reason": "Audio chat-template sampling-rate default vs Metal quantized batch generation corruption; different code paths." + "reason": "Audio chat-template sampling-rate bug vs Metal quantized batched generation corruption; different code paths." }, { - "left": "issue:43208", - "right": "issue:47013", + "left": "issue:43582", + "right": "issue:47331", "accept": false, - "reason": "Both are xLSTM-related, but one reports multiple training blockers for small models and the other is a specific return_last_states crash." + "reason": "Allocator warmup TypeError on Apple Silicon is unrelated to the Metal quantization row-corruption bug." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Both are Whisper token-handling issues, but one is batch_decode filtering and the other is empty-transcription generation failure." + "reason": "Both are Whisper-related, but one is skip_special_tokens handling in batch_decode and the other is empty-transcription failure after align_special_tokens." }, { - "left": "issue:43746", - "right": "issue:46326", + "left": "issue:43208", + "right": "issue:47013", "accept": false, - "reason": "PEFT local-checkpoint loading vs mistral_common continue_final_message behavior; different features and failures." + "reason": "Both are xLSTM training bugs, but they describe distinct crashes and should be tracked separately." }, { - "left": "issue:43317", - "right": "issue:43482", + "left": "issue:42915", + "right": "issue:45084", "accept": false, - "reason": "Different loading/offload contexts: dequantized device-map offload vs GGUF model loading." + "reason": "Different subsystems and failures: Qwen3Moe FP8 loading vs template compilation." }, { - "left": "issue:42915", - "right": "issue:45084", + "left": "issue:43792", + "right": "issue:46869", + "accept": false, + "reason": "Whisper can't-run issue vs YOLO multi-GPU device-map failure; unrelated." + }, + { + "left": "issue:43317", + "right": "issue:43482", "accept": false, - "reason": "Qwen3Moe FP8 config loading vs a template compilation error; unrelated." + "reason": "dequantized device_map offload failure vs GGUF model loading failure; different problems." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative about the heterogeneous cluster. The only accepted soft edge, issue:43452 \u2194 issue:43482, is reasonably supported by the titles as the same GGUF-loading regression, while the other pairwise rejections avoid over-merging distinct model/subsystem bugs." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3778,7 +3862,7 @@ "issue:43645", "issue:43646", "issue:43742", - "issue:43792", + "issue:43746", "issue:43819", "issue:43872", "issue:43994", @@ -3802,17 +3886,17 @@ "issue:45325", "issue:45923", "issue:45987", + "issue:46326", "issue:46661", "issue:46821", - "issue:46869", "issue:46899", "issue:47254", "issue:47448" ], "soft_pairs": [ - "issue:43792|issue:46869", - "issue:44403|issue:45081", + "issue:43746|issue:46326", "issue:43645|issue:46661", + "issue:44403|issue:45081", "issue:42915|issue:44291", "issue:44987|issue:45081", "issue:43646|issue:44534", @@ -3820,242 +3904,245 @@ "issue:45020|issue:47448", "issue:36032|issue:43742", "issue:43377|issue:43819", - "issue:43994|issue:44610", "issue:42915|issue:44589", + "issue:43994|issue:44610", + "issue:44008|issue:46899", "issue:28530|issue:34567", "issue:44062|issue:44987", "issue:44451|issue:45081", "issue:43531|issue:44877", - "issue:43404|issue:44466", "issue:44265|issue:44610", - "issue:44008|issue:46899", - "issue:44479|issue:45325", + "issue:43404|issue:44466", "issue:44849|issue:45325", - "issue:43322|issue:43540", + "issue:44479|issue:45325", "issue:43632|issue:46821", + "issue:43322|issue:43540", "issue:43089|issue:45987", "issue:43425|issue:44403", - "issue:45923|issue:47254", "issue:43296|issue:43531", + "issue:45923|issue:47254", "issue:44464|issue:45987", "issue:40444|issue:45325", "issue:43872|issue:44291" ], - "prepared_review_unit_hash": "fa14055c338cfc27a2856a5bd073b129aa7ac42b46b9f298ccd0d0b8c04b9e29", + "prepared_review_unit_hash": "206391d579b92f354bd1377993a3f836e9651aeb3751120c19ed54de8bf88587", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36857, - "estimated_input_tokens": 9215, - "estimated_eval_tokens": 18686 + "serialized_chars": 36965, + "estimated_input_tokens": 9242, + "estimated_eval_tokens": 18740 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose thematic grouping of mostly unrelated Transformers regressions and model-specific loading bugs. None of the soft pairs look like the same underlying defect, so I would not merge any of them as duplicates. If a representative issue is needed, the broad remote_code regression is the best umbrella candidate.", - "confidence": 0.89, - "canonical_issue_reason": "issue:45020 is the broadest representative of the cluster: it describes a general regression in recent Transformers versions affecting remote_code models, which is the closest thing here to an umbrella bug.", + "summary": "Most pairs are just thematically related transformer bugs, not true duplicates. The only clear duplicate-looking pair is the `_is_hf_initialized`/`init_empty_weights` regression affecting model loading, including the bitsandbytes case.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44291 is the clearest canonical report: it describes the underlying `init_empty_weights` / `_is_hf_initialized` breakage directly, while issue:43872 is a downstream bitsandbytes-specific manifestation of the same regression.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the most general and cluster-central issue among a set of otherwise distinct failures; it best represents the shared theme without pretending the rest are duplicates.", + "best_issue_reason": "issue:44291 is the best cluster representative because it names the shared root cause in a general way and is not tied to one downstream library or model variant.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43792", - "right": "issue:46869", + "left": "issue:43746", + "right": "issue:46326", "accept": false, - "reason": "Different models and failure modes: Whisper inference/runtime failure vs YOLO multi-GPU device-map loading." + "reason": "Different subsystems and failures: GraniteSpeech PEFT checkpoint loading vs Mistral-common tokenizer chat-template behavior." }, { - "left": "issue:44403", - "right": "issue:45081", + "left": "issue:43645", + "right": "issue:46661", "accept": false, - "reason": "One is noisy logging on load; the other is a tokenizer regex patch crash. Different code paths and symptoms." + "reason": "Both involve model initialization, but one is notebook/custom-model setup and the other is remote-code `config_class` lookup; not the same bug." }, { - "left": "issue:43645", - "right": "issue:46661", + "left": "issue:44403", + "right": "issue:45081", "accept": false, - "reason": "Both involve custom/remote-code models, but one is Jupyter/custom-model initialization and the other is AutoModel.from_pretrained missing config_class. Not the same bug." + "reason": "Unrelated symptoms: loading-log noise vs a Mistral tokenizer regex patch crash." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Both mention _is_hf_initialized, but one is FineGrainedFP8/Qwen3Moe quantization and the other is init_empty_weights loading. Distinct integrations." + "reason": "Both are model-loading failures, but one is Qwen3MoE FP8 config handling and the other is `_is_hf_initialized`/`init_empty_weights`; different code paths." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "Different model-loading failures: physical-intelligence/fast versus Mistral tokenizer regex patching." + "reason": "Different failures and model families; no clear shared underlying bug." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization breakage is unrelated to non-persistent buffer serialization junk." + "reason": "Custom model init regression vs non-persistent buffer serialization junk; too different to merge." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and MIMI batch-padding-mask mismatch are separate bugs." + "reason": "Both audio-related, but one is chat-template sampling-rate defaulting and the other is encoder padding-mask batching correctness." }, { "left": "issue:45020", "right": "issue:47448", "accept": false, - "reason": "Broad remote_code regressions are not the same as a specific arch-recognition failure for gemma4 12B." + "reason": "Generic remote-code breakage vs a specific model-architecture recognition problem; not the same concrete bug." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "A T5 tokenizer method-name conflict is unrelated to a MobileLLM load KeyError." + "reason": "Tokenizer constructor conflict vs loading a specific model failing with KeyError; unrelated." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "Both are audio-model bugs, but one is padding-mask batch consistency and the other is DAC STE/from_latents mismatch." + "reason": "Different audio-model bugs: batching/padding-mask mismatch vs DAC latent/forward mismatch without STE." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Different failure modes and components; no shared code-path bug." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Different vision bugs: SigLIP2 nonsensical outputs vs OmDet-Turbo processor/model input-size mismatch." + "reason": "Both are vision-model issues, but one is wrong outputs from AutoModel/pipeline and the other is a processor size mismatch." }, { - "left": "issue:42915", - "right": "issue:44589", + "left": "issue:44008", + "right": "issue:46899", "accept": false, - "reason": "Quantization/init issues versus a Float8 storage lookup failure are different code paths." + "reason": "Same general Gemma family, but the bugs differ: variable-name collision in forward vs unsafe dtype casting in multimodal embedders." }, { "left": "issue:28530", "right": "issue:34567", "accept": false, - "reason": "Early stopping metric selection and num_input_tokens_seen tracking are unrelated trainer behaviors." + "reason": "Training callback early-stopping logic vs token-count bookkeeping; unrelated." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "Tokenizer AddedToken kwarg duplication is unrelated to a model load failure for physical-intelligence/fast." + "reason": "Token constructor argument conflict vs loading a specific model with recent transformers; unrelated." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "Model-specific load failure for ScandiBERT is not the same as the Mistral regex-patching crash." + "reason": "Both are loading failures, but for different models and different tokenizer internals." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and granite_speech strict-config loading are different bugs." - }, - { - "left": "issue:43404", - "right": "issue:44466", - "accept": false, - "reason": "One is missing lm_head tying in a model class; the other is inconsistent serialization depending on device. Related area, but different defect." + "reason": "Qwen3-MoE sliding-window behavior vs strict config rejection for granite_speech; unrelated." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "torch.export failures with torch_compilable_check are unrelated to OmDet-Turbo input-resolution mismatch." + "reason": "torch.export compatibility issue vs processor/model input-size mismatch; not the same bug." }, { - "left": "issue:44008", - "right": "issue:46899", + "left": "issue:43404", + "right": "issue:44466", "accept": false, - "reason": "Different Gemma generations and different causes: variable-name collision vs unsafe dtype cast under 4-bit quantization." + "reason": "Both mention `lm_head.weight`, but one is missing weight tying in a model class and the other is device-dependent serialization; not the same code-path problem." }, { - "left": "issue:44479", + "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "A broad video-input regression across several Qwen models is not the same as the specific still-image rope-index scaling bug." + "reason": "Different Qwen2.5/Qwen3.5 behaviors: output_hidden_states bug vs temporal position-id scaling in Qwen2.5-VL." }, { - "left": "issue:44849", + "left": "issue:44479", "right": "issue:45325", "accept": false, - "reason": "Qwen3.5 hidden-state output bug and Qwen2.5-VL rope-index scaling are distinct model behaviors." + "reason": "Related Qwen VL area, but one is a broader video-input regression and the other is a still-image temporal position-id scaling bug." }, { - "left": "issue:43322", - "right": "issue:43540", + "left": "issue:43632", + "right": "issue:46821", "accept": false, - "reason": "A Llava Next segfault and a Qwen3OmniMoe video ValueError are different multimodal failures." + "reason": "Both are API regressions, but one is `_is_hf_initialized` and the other is `window_size`; unrelated." }, { - "left": "issue:43632", - "right": "issue:46821", + "left": "issue:43322", + "right": "issue:43540", "accept": false, - "reason": "_is_hf_initialized flag breakage is unrelated to TimesFM window_size AttributeError." + "reason": "Different multimodal models and failures: Llava Next load crash vs Qwen3OmniMoe video-processing ValueError." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation performance overhead is not the same as a StaticCache sequence-length return-type bug." + "reason": "Generation performance overhead vs cache length return-type correctness; not the same bug." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Torch version incompatibility and load-time noise are unrelated issues." + "reason": "Torch version compatibility vs loading-log noise; unrelated." }, { - "left": "issue:45923", - "right": "issue:47254", + "left": "issue:43296", + "right": "issue:43531", "accept": false, - "reason": "Missing supports_gradient_checkpointing on a remote-code model is different from a PEFT+DeepSpeed ZeRO-3 checkpointing error." + "reason": "PaddleOCR-VL/vLLM loading failure vs Qwen3-MoE sliding-window bug; no overlap." }, { - "left": "issue:43296", - "right": "issue:43531", + "left": "issue:45923", + "right": "issue:47254", "accept": false, - "reason": "PaddleOCR-VL load failure under vLLM is unrelated to Qwen3-MoE sliding_window behavior." + "reason": "Both mention gradient checkpointing, but one is a missing flag on a remote-code variant and the other is a checkpointing/ZeRO-3 runtime error; different issues." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency is not the same as StaticCache.get_seq_length returning a tensor." + "reason": "Chunked-generation inconsistency with compiled forward vs cache-length API contract bug; not the same concrete problem." }, { "left": "issue:40444", "right": "issue:45325", "accept": false, - "reason": "Multiple-image IterableDataset finetuning failure and still-image temporal position scaling are different Qwen2.5-VL bugs." + "reason": "Qwen2.5-VL dataset/multi-image finetuning failure vs Qwen2.5-VL still-image temporal position IDs; different parts of the stack." }, { "left": "issue:43872", "right": "issue:44291", - "accept": false, - "reason": "They share the _is_hf_initialized keyword theme, but one is bitsandbytes Int8Params construction and the other is init_empty_weights loading; too different to merge as one duplicate." + "accept": true, + "reason": "Same underlying regression: transformers v5\u2019s `_is_hf_initialized`/`init_empty_weights` change breaks loading, and bitsandbytes is just a downstream manifestation." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is well grounded in the packet. The canonical representative choice of issue:44291 over issue:43872 is supported by the titles, and the one accepted soft pair is a clear shared regression around `_is_hf_initialized` / `init_empty_weights`. The remaining soft-edge rejections stay conservative and do not overclaim duplicates." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -4081,7 +4168,6 @@ "issue:42796", "issue:43296", "issue:43381", - "issue:43421", "issue:43482", "issue:43525", "issue:43531", @@ -4093,21 +4179,22 @@ "issue:43883", "issue:44079", "issue:44220", + "issue:44265", "issue:44291", "issue:44387", "issue:44403", - "issue:44462", "issue:44488", "issue:44534", - "issue:44568", "issue:44589", "issue:44683", "issue:44756", "issue:44811", "issue:44991", + "issue:45072", "issue:45081", "issue:45216", "issue:45357", + "issue:45538", "issue:45542", "issue:45663", "issue:45812", @@ -4127,241 +4214,250 @@ "issue:43611|issue:44534", "issue:44991|issue:46315", "issue:44403|issue:44991", - "issue:45216|issue:45357", + "issue:44291|issue:44589", "issue:45812|issue:46491", + "issue:45216|issue:45357", "issue:42796|issue:43746", - "issue:44291|issue:44589", "issue:43883|issue:44488", "issue:43792|issue:44589", "issue:33666|issue:40444", - "issue:43381|issue:45923", "issue:43632|issue:44756", - "issue:40990|issue:44811", + "issue:43381|issue:45923", "issue:35707|issue:40444", - "issue:42796|issue:46710", + "issue:40990|issue:44811", "issue:43531|issue:44291", "issue:44079|issue:45663", + "issue:42796|issue:46710", "issue:46294|issue:47448", "issue:41720|issue:45081", "issue:43482|issue:43525", - "issue:44387|issue:47404", "issue:43296|issue:46821", + "issue:44387|issue:47404", "issue:42175|issue:43296", - "issue:42175|issue:45542", "issue:43883|issue:44991", - "issue:43421|issue:44568", + "issue:42175|issue:45542", "issue:44079|issue:44534", - "issue:44462|issue:45812" + "issue:43482|issue:45538", + "issue:44265|issue:45072" ], - "prepared_review_unit_hash": "45e251ae75ddffaebe0983dc91f1382a1c69c6af06d00fbf48ceb9644d6030bb", + "prepared_review_unit_hash": "edc051a6b9094068099b584f217c4627e5d528052bfe8f184c6464362422aefb", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36814, - "estimated_input_tokens": 9204, - "estimated_eval_tokens": 18664 + "serialized_chars": 36771, + "estimated_input_tokens": 9193, + "estimated_eval_tokens": 18642 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No true duplicate cluster here: the soft pairs mostly share vague vocabulary but describe different bugs, models, or subsystems. No PRs are present. If one issue must be used as a representative, the Granite tokenizer regression is the closest fit, but only weakly so.", - "confidence": 0.83, - "canonical_issue_reason": "Closest representative: issue:45812 is a broad, high-impact tokenizer regression affecting all Granite models and already has inbound references.", + "summary": "This cluster is mostly noisy similarity with unrelated issues, but two issue pairs look like true duplicates: Qwen3.5 save_pretrained key corruption (45216/45357) and TensorFlow backend packaging/install problems (42175/45542). No PRs are present.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 45216 is the best canonical anchor: it is the earliest, broadest report of the Qwen3.5 save_pretrained regression, and 45357 reads like a later follow-up of the same incorrect visual-encoder key serialization bug.", "canonical_pr_reason": null, - "best_issue_reason": "Best overall issue to anchor the cluster, if forced, because it is the broadest and most externally referenced regression among the tokenizer-related items.", + "best_issue_reason": "Issue 45216 is the strongest representative issue in the cluster because it names the concrete regression, has the clearest symptom description, and matches the later duplicate report most closely.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Different bugs: audio feature extraction failure vs compiled flex_attention breakage on torch >= 2.9." + "reason": "Both are torch-related failures, but they hit different code paths: fbank feature extraction vs compiled flex_attention on newer torch." }, { "left": "issue:43716", "right": "issue:46899", "accept": false, - "reason": "Both are multimodal dtype problems, but they affect different model families and code paths (Mistral-3 vs Gemma4)." + "reason": "Both mention dtype issues in multimodal models, but they are different models/components (Mistral-3 image preprocessing vs Gemma4 audio/vision embedders) with different failure modes." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "Related to Transformers v5 regressions, but one is loading with base_model_prefix and the other is non-persistent buffer initialization; not the same fix." + "reason": "Both are Transformers v5 regressions, but one is model loading with base_model_prefix and the other is non-persistent buffer initialization; not the same bug." }, { "left": "issue:44991", "right": "issue:46315", "accept": false, - "reason": "Tokenizer loading failure for a specific model vs convert_tokens_to_ids performance regression; different behaviors and remedies." + "reason": "Tokenizer-related, but one is a load failure for a specific tokenizer while the other is a performance regression in convert_tokens_to_ids." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Unnecessary logging/noise when loading a transformer is not the same as tokenizer loading failure for EMBEDDIA/est-roberta." + "reason": "One is about noisy loading logs, the other is a tokenizer loading failure; different symptoms and fixes." }, { - "left": "issue:45216", - "right": "issue:45357", + "left": "issue:44291", + "right": "issue:44589", "accept": false, - "reason": "Same model family and regression theme, but these are distinct save_pretrained visual-key bugs at different versions/symptoms." + "reason": "Both involve Transformers v5 compatibility, but they are distinct issues: unexpected _is_hf_initialized argument vs missing Float8 storage object." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Granite token IDs regression vs CodeLlama leading-space round-trip bug; both tokenizer bugs, but not the same underlying issue." + "reason": "Both are tokenizer regressions, but they affect different models and different behaviors (Granite token IDs vs CodeLlama leading-space round-trip)." }, { - "left": "issue:42796", - "right": "issue:43746", - "accept": false, - "reason": "Tokenizer BPE marker handling vs PEFT/local-checkpoint loading for GraniteSpeech; unrelated failure modes." + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same underlying Qwen3.5 save_pretrained regression: both report incorrect checkpoint serialization, with 45357 acting as a later reproduction/follow-up of the same visual-encoder key bug." }, { - "left": "issue:44291", - "right": "issue:44589", + "left": "issue:42796", + "right": "issue:43746", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized argument issue vs missing Float8 storage type; both v5-related but separate breakages." + "reason": "Tokenizer BPE-marker handling and GraniteSpeech PEFT checkpoint loading are unrelated bugs." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Model attribute error on Molmo vs a tokenizer loading report for cjvt/sleng-bert; not the same code path." + "reason": "Both are model loading failures, but one is a missing all_tied_weights_keys attribute and the other is a specific tokenizer/model load issue; not the same code path." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper model runtime failure vs Float8 storage lookup error; distinct runtime/serialization problems." + "reason": "Whisper run failure and Float8 storage lookup error are unrelated." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Qwen2-VL multi-GPU training question vs Qwen2.5-VL IterableDataset failure with multiple images; related family, different bug." - }, - { - "left": "issue:43381", - "right": "issue:45923", - "accept": false, - "reason": "Gradient checkpointing eval-mode error is unrelated to missing supports_gradient_checkpointing on a trust_remote_code variant." + "reason": "Both are Qwen2-VL-family training issues, but one is multi-GPU training and the other is iterable-dataset finetuning with multiple images per prompt; different failures." }, { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "_is_hf_initialized regression and mmap OOM on Strix Halo are separate v5 issues." + "reason": "Transformers v5 _is_hf_initialized breakage and Strix Halo mmap OOM are unrelated." }, { - "left": "issue:40990", - "right": "issue:44811", + "left": "issue:43381", + "right": "issue:45923", "accept": false, - "reason": "Perplexity regression on GPT-OSS is unrelated to Whisper processor.batch_decode ignoring skip_special_tokens." + "reason": "Both mention gradient checkpointing, but one is eval-mode usage and the other is a missing supports_gradient_checkpointing flag; distinct issues." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is a generation-cache issue, not a multimodal finetuning dataset bug." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multimodal finetuning with multiple images per prompt." }, { - "left": "issue:42796", - "right": "issue:46710", + "left": "issue:40990", + "right": "issue:44811", "accept": false, - "reason": "Tokenizer BPE marker handling and model accuracy regression on DeepSeek-R1 are different problems." + "reason": "GPT-OSS perplexity regression and Whisper batch_decode special-token handling are unrelated." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window bug and init_empty_weights/_is_hf_initialized TypeError are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights _is_hf_initialized argument are separate bugs." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment bug and Gemma4 FSDP2 shared_kv_states KeyError are different implementation defects." + "reason": "ModelOutput key assignment and Gemma4 FSDP2 shared_kv_states rebuild are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer BPE-marker handling and DeepSeek accuracy regression are different problems." }, { "left": "issue:46294", "right": "issue:47448", "accept": false, - "reason": "Video chat-template feature/token mismatch vs unrecognized gemma4 12B architecture; different areas entirely." + "reason": "Qwen3.5 video chat-template mismatch and Gemma4 architecture recognition are unrelated." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Qwen3 auto device map cudaErrorAssert on A800 is not the same as the Mistral regex patch crash." + "reason": "Qwen3 auto device mapping cudaErrorAssert and Mistral regex patch crashes are unrelated." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "Qwen2.5-GGUF loading failure with transformers v5 and missing pad_token_id on Llama4Config are distinct loading errors." + "reason": "GGUF loading failure in Transformers v5 and missing pad_token_id on Llama4Config are different incompatibilities." }, { - "left": "issue:44387", - "right": "issue:47404", + "left": "issue:43296", + "right": "issue:46821", "accept": false, - "reason": "Int4 quantization memory regression vs device_map loading OOM from conversion-op transients; related symptom, different root cause." + "reason": "PaddleOCR-VL loading and TimesFM window_size AttributeError are unrelated." }, { - "left": "issue:43296", - "right": "issue:46821", + "left": "issue:44387", + "right": "issue:47404", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM and TimesFM window_size AttributeError are unrelated model-specific bugs." + "reason": "Both are OOM-related, but one is int4 quantization memory growth and the other is device_map transient allocation during conversion; different root causes." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Missing TensorFlow backend under pip install '.[torch]' is unrelated to PaddleOCR-VL loading in vLLM." + "reason": "TensorFlow backend packaging and PaddleOCR-VL load failure are unrelated." }, { "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Molmo missing all_tied_weights_keys and tokenizer loading for est-roberta are separate failures." + "reason": "Missing all_tied_weights_keys and tokenizer loading for EMBEDDIA/est-roberta are unrelated." }, { - "left": "issue:43421", - "right": "issue:44568", - "accept": false, - "reason": "Runtime special-token post-processor update feature and mdeberta-v3 add_special_tokens BOS/EOS regression are not the same bug." + "left": "issue:42175", + "right": "issue:45542", + "accept": true, + "reason": "Both describe the same packaging/back-end dependency problem around TensorFlow being absent or incorrectly omitted while the tf backend is partially installed." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "ModelOutput key handling and non-persistent buffer junk fill are different internal-state bugs." + "reason": "ModelOutput key handling and non-persistent buffer junk initialization are unrelated." }, { - "left": "issue:44462", - "right": "issue:45812", + "left": "issue:43482", + "right": "issue:45538", "accept": false, - "reason": "Both are AutoTokenizer regressions, but one ignores tokenizer.json in a repo and the other produces wrong Granite token IDs; not mergeable as one fix." - } - ] - }, - "evaluator_result": null, - "evaluator_used": false, + "reason": "Qwen2.5-GGUF loading and CLIPTokenizer model_max_length behavior are unrelated." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export failures with torch_compilable_check and dtype mismatches in bfloat16 inference are different issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The two accepted duplicate-like pairs are well supported by the issue titles and descriptions, and the rest of the soft-edge rejections avoid overclaiming same-bug matches. No PRs are present, as stated." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -4381,41 +4477,41 @@ "issue:40444", "issue:40990", "issue:41720", - "issue:41950", "issue:42796", - "issue:43262", "issue:43317", "issue:43377", + "issue:43421", "issue:43454", - "issue:43482", "issue:43606", "issue:43611", "issue:43644", "issue:43646", - "issue:43746", "issue:43818", "issue:43828", "issue:43856", "issue:43873", "issue:43994", "issue:44079", + "issue:44186", "issue:44265", "issue:44291", + "issue:44462", "issue:44466", "issue:44479", "issue:44509", + "issue:44568", + "issue:44661", "issue:44829", "issue:44877", "issue:44898", - "issue:45072", "issue:45325", "issue:45357", "issue:45381", "issue:45479", "issue:45491", - "issue:45538", "issue:45584", "issue:45701", + "issue:45812", "issue:45823", "issue:45834", "issue:46402", @@ -4424,242 +4520,234 @@ "issue:46821", "issue:46909", "issue:47217", - "issue:47448", - "issue:47458" + "issue:47448" ], "soft_pairs": [ - "issue:43482|issue:45538", - "issue:44265|issue:45072", + "issue:44462|issue:45812", + "issue:46661|issue:47448", "issue:43317|issue:43856", - "issue:43611|issue:43646", "issue:45325|issue:45834", - "issue:46661|issue:47448", - "issue:44509|issue:46909", + "issue:43611|issue:43646", "issue:40444|issue:45381", + "issue:44509|issue:46909", "issue:44291|issue:44877", + "issue:43421|issue:44568", "issue:43644|issue:44079", "issue:43377|issue:45491", "issue:45325|issue:45381", "issue:43994|issue:44898", "issue:40444|issue:47217", - "issue:41950|issue:47448", "issue:43454|issue:44466", - "issue:45325|issue:46697", + "issue:44661|issue:47448", "issue:43873|issue:45701", - "issue:43818|issue:46402", - "issue:43818|issue:45357", + "issue:45325|issue:46697", "issue:43606|issue:44265", - "issue:42796|issue:45381", + "issue:43818|issue:45357", + "issue:43818|issue:46402", "issue:44479|issue:45823", + "issue:42796|issue:45381", "issue:43828|issue:46821", "issue:44829|issue:45479", - "issue:43262|issue:43746", + "issue:41720|issue:45823", "issue:40990|issue:45584", - "issue:43646|issue:47458", - "issue:41720|issue:45823" + "issue:44186|issue:44265" ], - "prepared_review_unit_hash": "3256f62b81659a33e563104434233cd90b9fc9a316d392eddd3e175070a17af3", + "prepared_review_unit_hash": "211454b1344a6bc4f3f113cc048e7817156c00a3dc1be83f8f6e2940eb277ccf", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36656, - "estimated_input_tokens": 9164, - "estimated_eval_tokens": 18584 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35851, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a false-positive cluster: the items span many unrelated bugs across different models, loaders, processors, docs, and training paths. None of the soft pairs looks like the same underlying bug/change, so I would not merge any of them.", - "confidence": 0.74, - "canonical_issue_reason": "No strong canonical issue stands out because the set is heterogeneous. If one must anchor the cluster, issue 44479 is the broadest concrete user-facing regression (Qwen video-input regression across multiple models), but it does not represent the whole set well.", + "summary": "This cluster is a loose collection of mostly unrelated Transformers issues: tokenizer regressions, multimodal/video position-id bugs, device-map/offload failures, model-loading regressions, and docs problems. I don't see any soft pair that is clearly the same underlying bug or change, so none should be merged as duplicates. If a single issue must represent the cluster, issue #43873 is the most active and best-documented bug report, but it is not a duplicate anchor for the others.", + "confidence": 0.95, + "canonical_issue_reason": "#43873 is the strongest triage anchor: it is open, highly discussed, has inbound references, and describes a concrete reproducible offloading/quantization failure. It is still distinct from the other issues, so this is only a representative choice, not a duplicate match.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44479 is the most broadly scoped, concrete regression in the list and the closest thing to a cluster anchor, though the overall cluster is still mixed.", + "best_issue_reason": "#43873 is the most suitable issue to keep as the representative item because it has the most activity and the clearest end-user impact/repro among this otherwise heterogeneous set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43482", - "right": "issue:45538", + "left": "issue:44462", + "right": "issue:45812", "accept": false, - "reason": "Different bug classes: GGUF/v5 model loading failure vs a tokenizer max_length/docs issue. No shared code path." + "reason": "Both involve AutoTokenizer/tokenization regressions, but one is about ignoring tokenizer.json from the repo while the other is about wrong token IDs for Granite models; different failures and likely different code paths." }, { - "left": "issue:44265", - "right": "issue:45072", + "left": "issue:46661", + "right": "issue:47448", "accept": false, - "reason": "`torch.export` with `torch_compilable_check` is a different failure mode from bfloat16 inference dtype mismatches." + "reason": "Both concern loading/recognizing remote or new model architectures, but one is an AutoModel config_class AttributeError and the other is an unrecognized Gemma4 arch; distinct loader failures." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "One is dequantized model loading/offload, the other is MoE training memory usage. Same broad area, but not the same bug." + "reason": "One is a device_map/offload load failure with quantization, the other is a training-memory issue in Qwen3 MoE; same broad area, not the same bug." }, { - "left": "issue:43611", - "right": "issue:43646", + "left": "issue:45325", + "right": "issue:45834", "accept": false, - "reason": "Both are v5 regressions, but one is about `base_model_prefix` loading and the other about custom model initialization; different concrete paths." + "reason": "Both are multimodal model bugs, but one is Qwen2.5-VL temporal position-id scaling and the other is a Kosmos2.5 OCR index error; different models and root causes." }, { - "left": "issue:45325", - "right": "issue:45834", + "left": "issue:43611", + "right": "issue:43646", "accept": false, - "reason": "Both involve model internals, but the titles point to different models and different failures (position-id scaling vs OCR index error)." + "reason": "Both mention Transformers 5.0 loading regressions, but one is base_model_prefix-related loading and the other is custom model initialization; not the same concrete issue." }, { - "left": "issue:46661", - "right": "issue:47448", + "left": "issue:40444", + "right": "issue:45381", "accept": false, - "reason": "Remote-code `config_class` AttributeError is unrelated to Gemma4 architecture recognition." + "reason": "Both are Qwen2.5-VL multimodal regressions, but one is IterableDataset with multiple images per prompt and the other is video vision_position_ids; different input shapes and failure modes." }, { "left": "issue:44509", "right": "issue:46909", "accept": false, - "reason": "Both are docs-related, but they concern unrelated documentation problems and different endpoints." + "reason": "Both are documentation issues, but one is about removed pipeline task docs and the other is a broken DIA example; unrelated doc problems." }, { - "left": "issue:40444", - "right": "issue:45381", + "left": "issue:44291", + "right": "issue:44877", "accept": false, - "reason": "Both are Qwen2.5-VL vision-position issues, but one is multi-image finetuning with IterableDataset and the other is video input; not clearly the same bug." + "reason": "One is init_empty_weights/_is_hf_initialized loading incompatibility, the other is strict config rejecting granite_speech; separate regressions." }, { - "left": "issue:44291", - "right": "issue:44877", + "left": "issue:43421", + "right": "issue:44568", "accept": false, - "reason": "`init_empty_weights` argument mismatch is different from strict config blocking `granite_speech` loading." + "reason": "Both involve special tokens, but one is runtime post-processor updates in TokenizersBackend and the other is add_special_tokens behavior for a specific tokenizer; not the same bug." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Non-persistent buffer junk and `ModelOutput` key assignment are unrelated internals." + "reason": "Non-persistent buffer serialization junk and ModelOutput key assignment are different internal data-structure bugs." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Both mention batching/padding, but they affect different models and different mechanisms (padding mask support vs sliding-window/all-padding NaNs)." + "reason": "Both are batching/padding-related, but one is MIMI encoder missing padding-mask support and the other is Gemma3 NaNs from mixed-length sliding-window attention; different mechanisms." }, { "left": "issue:45325", "right": "issue:45381", "accept": false, - "reason": "Same model family, but one is still-image temporal scaling and the other is video input vision-position IDs; not the same concrete defect." + "reason": "Both are Qwen2.5-VL position-id issues, but one is still-image temporal scaling and the other is video-input vision_position_ids; related theme, different defect." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "Different models and APIs: SigLIP2 AutoModel/pipeline output issues vs Perceiver interpolation padding behavior." + "reason": "One is a SigLIP2 AutoModel/pipeline output quality problem, the other is a Perceiver interpolate_pos_encoding failure; not the same bug." }, { "left": "issue:40444", "right": "issue:47217", "accept": false, - "reason": "Both involve VL image handling, but one is finetuning with multiple images per prompt and the other is image placeholders in user text; too different to merge." + "reason": "Both involve VL image placeholder handling, but one is finetuning with multiple images per prompt and the other is failure when the placeholder appears in user text; distinct scenarios." }, { - "left": "issue:41950", + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both are tied-weight/lm_head-related, but one is AyaVision weight tying and the other is inconsistent lm_head serialization by device; different code paths." + }, + { + "left": "issue:44661", "right": "issue:47448", "accept": false, - "reason": "Video-classification pipeline processor lookup is unrelated to Gemma4 architecture recognition." + "reason": "Both are model/arch recognition issues, but one is add-new-model-like failing inside TOKENIZER_MAPPING_NAMES and the other is Gemma4 arch detection; unrelated loader logic." }, { - "left": "issue:43454", - "right": "issue:44466", + "left": "issue:43873", + "right": "issue:45701", "accept": false, - "reason": "One is failure to tie `lm_head` weights causing bad generation; the other is inconsistent serialization of `lm_head.weight`. Related theme, but different bug." + "reason": "Quantization/offload behavior and tokenization changing across versions are unrelated regressions." }, { "left": "issue:45325", "right": "issue:46697", "accept": false, - "reason": "Qwen2.5-VL position-id scaling and Swin2SR over-padding are unrelated image-processing bugs." + "reason": "Qwen2.5-VL position-id scaling and Swin2SR over-padding are unrelated preprocessing/model bugs." }, { - "left": "issue:43873", - "right": "issue:45701", + "left": "issue:43606", + "right": "issue:44265", "accept": false, - "reason": "Quantization/offload behavior and tokenization changes are completely different issues." + "reason": "CPU offload device mismatch for bark-small and torch.export failures with torch_compilable_check are separate backend issues." }, { "left": "issue:43818", - "right": "issue:46402", + "right": "issue:45357", "accept": false, - "reason": "Video-LLaVA temporal-attention/weight-sharing problem is unrelated to Sam3Model text-encoder loading." + "reason": "Video-LLaVA temporal-attention/weight-sharing bug and Qwen3.5 save_pretrained visual-key regression are different model-specific problems." }, { "left": "issue:43818", - "right": "issue:45357", + "right": "issue:46402", "accept": false, - "reason": "One is a Video-LLaVA architecture bug, the other is incorrect saved visual encoder keys in Qwen3.5." + "reason": "Video-LLaVA architecture/weights issue and Sam3Model text encoder weight loading are unrelated." }, { - "left": "issue:43606", - "right": "issue:44265", + "left": "issue:44479", + "right": "issue:45823", "accept": false, - "reason": "CPU offload device mismatch is a different code path from `torch.export`/compilable-check failures." + "reason": "Both are device/version regressions in multimodal models, but one is Qwen2.5-VL video input and the other is Gemma4 PLE device mismatch; not the same concrete failure." }, { "left": "issue:42796", "right": "issue:45381", "accept": false, - "reason": "Tokenizer BPE marker handling has nothing to do with Qwen2.5-VL video vision-position IDs." - }, - { - "left": "issue:44479", - "right": "issue:45823", - "accept": false, - "reason": "Qwen video-input regression and Gemma4 PLE device mismatch are different models and different failures." + "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video vision_position_ids are unrelated bugs." }, { "left": "issue:43828", "right": "issue:46821", "accept": false, - "reason": "Autocast dtype mismatch is unrelated to a missing `window_size` attribute." + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and TimesFM window_size AttributeError are completely different issues." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Both involve degenerate training behavior, but one is tied to Flash Attention 3 and the other to `num_labels=1` classification setup; not the same bug." + "reason": "Both describe degenerate training outcomes, but one is tied to flash_attention_3 and the other to single_label_classification with num_labels=1; separate causes." }, { - "left": "issue:43262", - "right": "issue:43746", + "left": "issue:41720", + "right": "issue:45823", "accept": false, - "reason": "Audio chat-template sample-rate defaulting and PEFT adapter loading are unrelated." + "reason": "Both are device_map='auto' failures, but one is a cudaErrorAssert on A800 for Qwen3 and the other is a device mismatch in Gemma4 PLE; distinct bugs." }, { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "High perplexity on WikiText and Whisper empty-transcription generation failure are unrelated model behaviors." - }, - { - "left": "issue:43646", - "right": "issue:47458", - "accept": false, - "reason": "Custom model initialization breakage and GPT2 `c_proj` weight initialization are different issues." + "reason": "High perplexity on gpt-oss-20b and Whisper empty-transcription failure after align_special_tokens are unrelated model-specific issues." }, { - "left": "issue:41720", - "right": "issue:45823", + "left": "issue:44186", + "right": "issue:44265", "accept": false, - "reason": "Both involve `device_map=auto`, but one is a CUDA assert on A800 and the other is a forward-time device mismatch in Gemma4; different failures." + "reason": "LayoutLMv2Tokenizer NER/padding crashes and torch.export failures with torch_compilable_check are unrelated." } ] }, @@ -4694,20 +4782,20 @@ "issue:42915", "issue:42947", "issue:43257", + "issue:43262", "issue:43317", "issue:43322", "issue:43377", - "issue:43450", "issue:43454", "issue:43550", "issue:43572", "issue:43606", + "issue:43646", "issue:43720", + "issue:43746", "issue:43749", "issue:43819", "issue:43873", - "issue:44186", - "issue:44265", "issue:44387", "issue:44534", "issue:44568", @@ -4729,18 +4817,20 @@ "issue:46315", "issue:46821", "issue:47404", - "issue:47405" + "issue:47405", + "issue:47458" ], "soft_pairs": [ - "issue:44186|issue:44265", "issue:34634|issue:35707", "issue:35707|issue:41720", + "issue:43262|issue:43746", + "issue:43646|issue:47458", "issue:43606|issue:43819", - "issue:43454|issue:43550", "issue:45823|issue:46821", "issue:42617|issue:42915", - "issue:43257|issue:45923", + "issue:43454|issue:43550", "issue:40990|issue:42947", + "issue:43257|issue:45923", "issue:43454|issue:45072", "issue:47404|issue:47405", "issue:44568|issue:45920", @@ -4751,233 +4841,232 @@ "issue:36331|issue:42915", "issue:41720|issue:43322", "issue:44877|issue:44991", - "issue:43454|issue:43720", "issue:43749|issue:44534", + "issue:43454|issue:43720", "issue:43317|issue:43873", "issue:43377|issue:44898", - "issue:41720|issue:44977", "issue:41720|issue:42617", + "issue:41720|issue:44977", "issue:45325|issue:46315", "issue:44387|issue:45704", - "issue:43572|issue:45759", - "issue:43322|issue:46050", "issue:40990|issue:45701", - "issue:43450|issue:45701" + "issue:43322|issue:46050", + "issue:43572|issue:45759" ], - "prepared_review_unit_hash": "7f45326dec9644674d0e8cfe64b2c7d5f4d23ce82da7b03edcdcef16d68109e3", + "prepared_review_unit_hash": "c013198d83dada1b4701d94bab6ec91d0eae2e44c1db06ad27badac470a462b1", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36515, - "estimated_input_tokens": 9129, - "estimated_eval_tokens": 18514 + "serialized_chars": 37116, + "estimated_input_tokens": 9279, + "estimated_eval_tokens": 18814 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a set of lookalike but unrelated bug reports. The strongest common thread is offloading/loading/quantization regressions, but even those usually hit different code paths, models, or failure modes. I did not find a true duplicate pair worth merging.", - "confidence": 0.88, - "canonical_issue_reason": "No single issue cleanly dominates the cluster; issue 43873 is the broadest representative of the only coherent subgroup (offloading + quantization/loading regressions).", + "summary": "These items are mostly unrelated bug reports spanning tokenization, offloading, quantization, configs, and model-specific failures. I do not see any soft pair that is clearly the same underlying issue, so I reject all soft edges. There are no PRs in the set; if a single issue must anchor the cluster, the broadest/highest-activity report is 43873.", + "confidence": 0.44, + "canonical_issue_reason": "Issue 43873 is the broadest and most active report here (offloading/quantization, open, multiple inbound refs), so it is the best cluster anchor even though the set is not a true duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the most general and reusable anchor for the offload/quantization subset, and it is open, making it the best representative if one must be chosen.", + "best_issue_reason": "43873 has the widest apparent scope and strongest discussion/ref signal, making it the most plausible representative issue among otherwise distinct reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44186", - "right": "issue:44265", - "accept": false, - "reason": "Tokenizer crash on NER/batching is a different bug class from torch.export failing on torch_compilable_check." - }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice_preset and progressive generation with inputs_embeds/past_key_values are unrelated code paths." + "reason": "Different subsystems and failure modes: BarkProcessor voice preset vs generation with inputs_embeds/past_key_values." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Generation cache/inputs_embeds behavior is not the same as auto device mapping causing cudaErrorAssert." + "reason": "Unrelated generation-cache bug vs Qwen3 device-mapping cuda assert." }, { - "left": "issue:43606", - "right": "issue:43819", + "left": "issue:43262", + "right": "issue:43746", "accept": false, - "reason": "Both mention mismatches, but one is CPU offload device placement and the other is DAC latent/STE behavior; different bugs." + "reason": "Audio chat-template sampling-rate default vs GraniteSpeech PEFT checkpoint loading; different code paths." }, { - "left": "issue:43454", - "right": "issue:43550", + "left": "issue:43646", + "right": "issue:47458", + "accept": false, + "reason": "General Transformers 5.0 custom init breakage vs GPT-2 c_proj initialization bug; not the same defect." + }, + { + "left": "issue:43606", + "right": "issue:43819", "accept": false, - "reason": "AyaVision weight tying and Bamba torch.compile SDPA failures are separate model-specific issues." + "reason": "Both involve model behavior, but one is Bark CPU offload device mismatch and the other is DAC latents/STE mismatch." }, { "left": "issue:45823", "right": "issue:46821", "accept": false, - "reason": "Gemma4 device mismatch during forward is unrelated to TimesFM's window_size AttributeError." + "reason": "Gemma4 device-mapped forward mismatch vs TimesFM window_size AttributeError; distinct components." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "A parallel script runtime problem is not the same as Qwen3Moe failing under FineGrainedFP8Config." + "reason": "3d_parallel.py runtime failure vs Qwen3Moe FP8 config failure; no shared concrete bug." }, { - "left": "issue:43257", - "right": "issue:45923", + "left": "issue:43454", + "right": "issue:43550", "accept": false, - "reason": "Missing gradient checkpointing flag and MOE weight conversion under accelerate+deepspeed are different failure modes." + "reason": "AyaVision lm_head tying bug vs Bamba torch.compile/SDPA issue; different model and root cause." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "High perplexity on a model is not the same bug as ineffective gradient checkpointing with LoRA." + "reason": "Perplexity regression on gpt-oss vs gradient checkpointing ineffective with LoRA; unrelated symptoms." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "MOE weight conversion with accelerate+deepspeed vs missing supports_gradient_checkpointing flag; separate issues." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "lm_head tying in AyaVision is unrelated to dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "AyaVision weight-tying bug vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:47404", "right": "issue:47405", "accept": false, - "reason": "One is OOM from transient allocation during device_map loading; the other is silent wrong-class conversion skipping leading to bad weights." + "reason": "Both are loading-related, but one is OOM during conversion transients and the other is silent wrong-Auto-class conversion skip." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is missing BOS/EOS insertion for a specific tokenizer while the other is wrong token IDs across multiple models." + "reason": "Both are tokenizer regressions, but one is missing BOS/EOS for a specific tokenizer and the other is wrong token IDs across several models." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "Missing padding-mask support in MIMI encoder is unrelated to OmDet-Turbo's processor/model input-size mismatch." + "reason": "MIMI padding-mask batching bug vs OmDet-Turbo processor image-size mismatch; different modalities and fixes." }, { "left": "issue:42796", "right": "issue:46209", "accept": false, - "reason": "BPE marker handling in a tokenizer is not the same as FP8 inference failure on B200." + "reason": "Tokenizer BPE-marker handling vs FP8 inference failure on B200; no shared code path." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Trainer compute_loss signature mismatch is unrelated to a 3d_parallel.py execution failure." + "reason": "Trainer loss signature regression vs parallel script execution failure; unrelated." }, { "left": "issue:36331", "right": "issue:39921", "accept": false, - "reason": "Unexpected compute_loss kwargs and Gemma3N projection errors are different bugs." + "reason": "Unexpected trainer kwarg vs Gemma3N special-token projection error; different layers of the stack." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "Trainer API breakage is not the same as Qwen3Moe/FP8 config failure." + "reason": "Trainer API mismatch vs Qwen3Moe FP8 config failure; no overlap beyond being runtime errors." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "cudaErrorAssert under auto device mapping is a different issue from Llava Next segfaults on load." + "reason": "Qwen3 device-map cuda assert vs Llava Next segmentation fault; different models and failure mechanisms." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config loading and tokenizer loading failures are separate components and failure paths." + "reason": "Strict config loading error vs tokenizer loading regression; not the same bug." }, { - "left": "issue:43454", - "right": "issue:43720", + "left": "issue:43749", + "right": "issue:44534", "accept": false, - "reason": "Weight tying in AyaVision is unrelated to packed-weight unpacking during accelerate loading." + "reason": "FSDP CPU RAM efficient loading breakage vs non-persistent buffers being filled with junk; both v5 regressions but distinct." }, { - "left": "issue:43749", - "right": "issue:44534", + "left": "issue:43454", + "right": "issue:43720", "accept": false, - "reason": "FSDP CPU RAM efficient loading and junk-filled non-persistent buffers are not the same bug." + "reason": "AyaVision weight tying issue vs BitNet packed-weight unpacking during accelerate load; separate model-loading paths." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Both are offload/quantization-related, but one is dequantized-model loading on gpu+cpu offload and the other is a broad offloading complaint; not clearly the same underlying bug." + "reason": "Both touch offloading, but one is dequantized-model loading on gpu+cpu offload and the other is broad quantized offloading behavior; too different to merge." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "Batch padding-mask behavior in MIMI encoder is unrelated to Perceiver resolution/interpolation issues." + "reason": "MIMI encoder padding-mask mismatch vs Perceiver non-default resolution interpolation failure; unrelated architecture and bug." }, { "left": "issue:41720", - "right": "issue:44977", + "right": "issue:42617", "accept": false, - "reason": "Device-mapping assert and flash-attention generation problems are different code paths." + "reason": "Device-mapping assert on Qwen3 vs failing 3d_parallel.py; no common underlying defect." }, { "left": "issue:41720", - "right": "issue:42617", + "right": "issue:44977", "accept": false, - "reason": "Auto device mapping failure is not the same as a 3d_parallel.py runtime issue." + "reason": "Same model family only at a high level; flash-attention generation issue is not the same as auto device-map cuda assert." }, { "left": "issue:45325", "right": "issue:46315", "accept": false, - "reason": "Temporal position-id scaling in Qwen2.5-VL is unrelated to tokenizer lookup performance regression." + "reason": "Qwen2.5-VL rope/position-id scaling bug vs slow convert_tokens_to_ids regression; different tokenizer/multimodal code paths." }, { "left": "issue:44387", "right": "issue:45704", "accept": false, - "reason": "Int4 quantization OOM and an apex RMSNorm memory leak are different memory bugs." + "reason": "Int4 quantization memory regression vs apex FusedRMSNorm memory leak; both memory-related but not the same bug." }, { - "left": "issue:43572", - "right": "issue:45759", + "left": "issue:40990", + "right": "issue:45701", "accept": false, - "reason": "Missing pad_token_idx in a config is unrelated to AutoModelForCausalLM not unwrapping text_config." + "reason": "High perplexity regression vs version-dependent tokenization change; not enough to treat as one defect." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "Llava Next loading segfaults and sinq save/load breakage are different failure modes." - }, - { - "left": "issue:40990", - "right": "issue:45701", - "accept": false, - "reason": "Perplexity regression on gpt-oss and version-dependent tokenization changes do not match." + "reason": "Segfault loading Llava Next vs quantized model save/load breakage; different symptoms and models." }, { - "left": "issue:43450", - "right": "issue:45701", + "left": "issue:43572", + "right": "issue:45759", "accept": false, - "reason": "Video processor shape bugs are unrelated to tokenization changes across versions." + "reason": "Missing pad_token_idx in StableLmConfig vs unwrap text_config for composite Qwen multimodal configs; unrelated config issues." } ] }, @@ -5011,6 +5100,7 @@ "issue:43295", "issue:43317", "issue:43404", + "issue:43450", "issue:43531", "issue:43550", "issue:43572", @@ -5019,7 +5109,6 @@ "issue:43606", "issue:43646", "issue:43708", - "issue:43716", "issue:43720", "issue:43873", "issue:44008", @@ -5051,231 +5140,231 @@ "issue:47404" ], "soft_pairs": [ + "issue:43450|issue:45701", "issue:41553|issue:43582", - "issue:43873|issue:44756", "issue:45685|issue:46821", - "issue:44466|issue:45357", + "issue:43873|issue:44756", "issue:44186|issue:45072", + "issue:44466|issue:45357", "issue:43295|issue:45701", "issue:26413|issue:28530", "issue:44451|issue:47020", "issue:43065|issue:46402", - "issue:43550|issue:43720", "issue:43550|issue:43606", - "issue:46612|issue:46858", + "issue:43550|issue:43720", "issue:43577|issue:43708", - "issue:44186|issue:44610", "issue:46821|issue:46869", - "issue:43646|issue:46665", - "issue:46897|issue:47317", + "issue:44186|issue:44610", + "issue:46612|issue:46858", + "issue:44008|issue:45823", "issue:41720|issue:45161", - "issue:42915|issue:43317", "issue:44423|issue:45406", - "issue:44008|issue:45823", + "issue:43646|issue:46665", + "issue:42915|issue:43317", + "issue:46897|issue:47317", "issue:44610|issue:44898", - "issue:28530|issue:33666", "issue:42915|issue:46291", + "issue:28530|issue:33666", "issue:43404|issue:45357", "issue:45704|issue:47404", - "issue:43531|issue:43572", - "issue:43716|issue:46821" + "issue:43531|issue:43572" ], - "prepared_review_unit_hash": "5efea9dc42f92e07ea9b7454aed9a2522c9969c8f4e6ce6d07d718154350305d", + "prepared_review_unit_hash": "ffc211e0c6f763fb8feb9b1d991149732cf3039e3084be094cccfb913b1204cc", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36229, - "estimated_input_tokens": 9058, - "estimated_eval_tokens": 18372 + "serialized_chars": 36226, + "estimated_input_tokens": 9057, + "estimated_eval_tokens": 18370 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are all issue reports, not pull requests. The pairings are mostly broad-theme similarities rather than the same underlying bug, so none of the soft edges should be merged as duplicates. Issue #43873 is the best representative of the cluster because it is the most active and broadly scoped loading/offload problem.", - "confidence": 0.84, - "canonical_issue_reason": "#43873 is the strongest cluster representative: it has the highest discussion/inbound-reference activity and describes a broadly impactful loading/quantization/offload failure rather than a narrowly model-specific edge case.", + "summary": "This cluster is a loose topical grouping, not a duplicate set: the items cover many unrelated bugs across trainer resume logic, tokenizers/processors, multimodal serving, device mapping/offload, quantization, and serialization. None of the proposed soft edges look like the same concrete underlying bug/change, so I would not merge any of them.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical duplicate stands out; the items span distinct bugs and subsystems rather than one shared underlying issue.", "canonical_pr_reason": null, - "best_issue_reason": "#43873 is the most suitable global issue anchor for triage because it is open, active, and centers on a general offloading/quantization failure mode with downstream references.", + "best_issue_reason": "issue:43708 is the strongest standalone report: it is specific, reproducible, and describes a concrete Trainer regression in checkpoint resume/max_steps calculation.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:41553", - "right": "issue:43582", + "left": "issue:43450", + "right": "issue:45701", "accept": false, - "reason": "Different subsystems and failures: tokenizer error messaging for Voxtral vs Apple Silicon allocator warmup TypeError." + "reason": "Different problems: batched video processor output shape vs version-dependent tokenization behavior." }, { - "left": "issue:43873", - "right": "issue:44756", + "left": "issue:41553", + "right": "issue:43582", "accept": false, - "reason": "Both involve memory/offload concerns, but one is quantization offloading behavior and the other is mmap OOM on Strix Halo; not the same bug." + "reason": "Unrelated: bad AutoTokenizer error for Voxtral vs Apple Silicon TypeError in caching_allocator_warmup." }, { "left": "issue:45685", "right": "issue:46821", "accept": false, - "reason": "Unrelated issues: MPS histogram kernel type support vs TimesFM window_size attribute handling." + "reason": "Different components and failures: MPS MoE histogram issue vs TimesFM window_size AttributeError." }, { - "left": "issue:44466", - "right": "issue:45357", + "left": "issue:43873", + "right": "issue:44756", "accept": false, - "reason": "Both concern save/load serialization, but one is tied lm_head serialization and the other is incorrect visual encoder key saving in Qwen3.5." + "reason": "Both involve memory/offload, but the concrete bugs differ: quantization offload behavior vs mmap OOM on Strix Halo." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Tokenizer crash on NER/batched padding is unrelated to dtype mismatches in bfloat16 inference." + "reason": "Tokenizer crash on NER/batched padding is unrelated to bfloat16 dtype mismatches in model inference." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both are serialization-related, but one is lm_head weight tying and the other is visual encoder key saving." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "Both mention version-related regressions, but they affect different APIs and behaviors; no shared concrete bug path." + "reason": "Regression around processor.tokenizer/images is not the same as generic tokenization changes across versions." }, { "left": "issue:26413", "right": "issue:28530", "accept": false, - "reason": "Trainer checkpoint resumption failure is unrelated to early stopping metric resolution/disablement." + "reason": "Different Trainer issues: resume_from_checkpoint epoch_iterator failure vs early stopping metric lookup/disablement." }, { "left": "issue:44451", "right": "issue:47020", "accept": false, - "reason": "Model loading failure for ScandiBERT is not the same as XLMRobertaTokenizer passing the wrong type to Unigram." + "reason": "Different tokenizer loading bugs for different models and code paths." }, { "left": "issue:43065", "right": "issue:46402", "accept": false, - "reason": "Same model family, but different components: dummy Conv2d in Sam3PixelDecoder vs missing text encoder weights in Sam3Model." - }, - { - "left": "issue:43550", - "right": "issue:43720", - "accept": false, - "reason": "Both are model/runtime failures, but one is torch.compile+SDPA on Bamba and the other is packed-weight unpacking during accelerate loading for BitNet." + "reason": "Different SAM3 bugs: dummy Conv2d in pixel decoder vs text encoder weights not loading." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "Different concrete failures: torch.compile/SDPA on Bamba vs CPU offload device mismatch on bark-small." + "reason": "Different model/runtime failures: Bamba torch.compile+SDPA vs Bark CPU offload device mismatch." }, { - "left": "issue:46612", - "right": "issue:46858", + "left": "issue:43550", + "right": "issue:43720", "accept": false, - "reason": "Both touch generation/cache logic, but one is beam-search cache reorder omission and the other is static-cache seq-length tensor arithmetic under torch.compile." + "reason": "Different concrete causes: SDPA/compile failure vs packed weights not unpacked during accelerate loading." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Blip2 dtype propagation bug is unrelated to Trainer max_steps recalculation when resuming with a different batch size." + "reason": "Blip2 dtype propagation and Trainer checkpoint-resume max_steps are unrelated bugs." }, { - "left": "issue:44186", - "right": "issue:44610", + "left": "issue:46821", + "right": "issue:46869", "accept": false, - "reason": "Tokenizer padding/truncation crash is unrelated to OmDet-Turbo producing the wrong image resolution." + "reason": "TimesFM window_size AttributeError is unrelated to YOLO multi-GPU device_map auto loading." }, { - "left": "issue:46821", - "right": "issue:46869", + "left": "issue:44186", + "right": "issue:44610", "accept": false, - "reason": "TimesFM window_size attribute error is unrelated to YOLO multi-GPU device-map loading failures." + "reason": "Different vision/text preprocessing bugs with different root causes and affected models." }, { - "left": "issue:43646", - "right": "issue:46665", + "left": "issue:46612", + "right": "issue:46858", "accept": false, - "reason": "Custom model initialization breakage and from_pretrained behavior inside Python classes are related only at a high level, not the same bug." + "reason": "Both concern generation/cache internals, but one is missing beam cache reorder support and the other is a device-tensor arithmetic issue in static cache prefill." }, { - "left": "issue:46897", - "right": "issue:47317", + "left": "issue:44008", + "right": "issue:45823", "accept": false, - "reason": "Both are training-loss bugs, but they involve different models and no evidence of the same computation path or fix." + "reason": "Different multimodal model bugs: Gemma3n forward variable collision vs Gemma4 PLE device mismatch." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Qwen3 auto device-mapping cudaErrorAssert is not the same as GPT-OSS MoE tensor-parallel support failing." + "reason": "Distinct parallelism/device-placement failures in different models." }, { - "left": "issue:42915", - "right": "issue:43317", + "left": "issue:44423", + "right": "issue:45406", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and dequantized model loading with cpu offload are distinct quantization/loading problems." + "reason": "Both involve serve/processor paths, but one is a string-to() crash in continuous batching and the other is a missing _tokenizer attribute." }, { - "left": "issue:44423", - "right": "issue:45406", + "left": "issue:43646", + "right": "issue:46665", "accept": false, - "reason": "Both are serve-time multimodal crashes, but they fail in different processors/code paths with different missing attributes." + "reason": "Related to model loading/init, but not the same concrete bug or fix." }, { - "left": "issue:44008", - "right": "issue:45823", + "left": "issue:42915", + "right": "issue:43317", "accept": false, - "reason": "Gemma 3n forward variable-name collision is unrelated to Gemma4 PLE device mismatch during forward." + "reason": "Different loading/offload failures: FineGrainedFP8Config on Qwen3Moe vs dequantized model loading with gpu+cpu offload." }, { - "left": "issue:44610", - "right": "issue:44898", + "left": "issue:46897", + "right": "issue:47317", "accept": false, - "reason": "Both are image preprocessing/resolution issues, but they affect different models and different mismatch mechanisms." + "reason": "Same broad symptom category, but different models and likely different loss-shift causes." }, { - "left": "issue:28530", - "right": "issue:33666", + "left": "issue:44610", + "right": "issue:44898", "accept": false, - "reason": "Early stopping metric lookup and Qwen2-VL multi-GPU training are unrelated." + "reason": "Different image-size preprocessing bugs in different model families." }, { "left": "issue:42915", "right": "issue:46291", "accept": false, - "reason": "FineGrainedFP8Config model failure is unrelated to kernels package import/version requirement errors." + "reason": "FP8 model loading and kernels package version/revision import failure are unrelated." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early stopping behavior is unrelated to Qwen2-VL multi-GPU training." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "Different model families and bugs: untied lm_head weights in Mistral3 vs incorrect visual encoder keys in Qwen3.5 save_pretrained." + "reason": "Different weight-serialization problems affecting different model components." }, { "left": "issue:45704", "right": "issue:47404", "accept": false, - "reason": "Both can lead to memory issues, but one is an apex RMSNorm memory leak and the other is device-map OOM from conversion-op transients." + "reason": "Both are memory-related, but the causes differ: apex fused RMSNorm leak vs unbudgeted contiguous() OOM during device_map loading." }, { "left": "issue:43531", "right": "issue:43572", "accept": false, - "reason": "Sliding-window handling in Qwen3-MoE is unrelated to missing pad_token_idx in StableLmConfig." - }, - { - "left": "issue:43716", - "right": "issue:46821", - "accept": false, - "reason": "Mistral-3 image preprocessor dtype mismatch is unrelated to TimesFM window_size attribute errors." + "reason": "Different issues: Qwen3-MoE sliding_window behavior vs StableLmConfig pad_token_idx regression." } ] }, @@ -5316,9 +5405,8 @@ "issue:43531", "issue:43540", "issue:43550", - "issue:43606", "issue:43653", - "issue:43720", + "issue:43716", "issue:43746", "issue:43782", "issue:43873", @@ -5340,6 +5428,7 @@ "issue:45704", "issue:45759", "issue:46665", + "issue:46821", "issue:46858", "issue:46894", "issue:47020", @@ -5348,210 +5437,217 @@ "issue:47337" ], "soft_pairs": [ - "issue:43441|issue:47337", + "issue:43716|issue:46821", "issue:28530|issue:30064", - "issue:44186|issue:44361", + "issue:43441|issue:47337", "issue:43881|issue:44488", - "issue:44734|issue:46858", + "issue:44186|issue:44361", "issue:44464|issue:46858", + "issue:44734|issue:46858", "issue:30064|issue:34634", - "issue:43452|issue:47020", - "issue:44671|issue:44743", "issue:29159|issue:36032", + "issue:43452|issue:47020", "issue:42915|issue:43322", + "issue:44671|issue:44743", "issue:44387|issue:45701", - "issue:45479|issue:47317", "issue:43873|issue:45704", - "issue:33357|issue:35707", "issue:43531|issue:45759", + "issue:33357|issue:35707", + "issue:45479|issue:47317", "issue:43540|issue:43782", "issue:43316|issue:44913", "issue:43550|issue:43653", - "issue:41720|issue:43746", "issue:45632|issue:45684", "issue:44534|issue:46665", + "issue:41720|issue:43746", "issue:46894|issue:47317", "issue:43257|issue:47254", - "issue:43606|issue:43720" + "issue:34634|issue:41720" ], - "prepared_review_unit_hash": "f6486226195d5bf14f7a45631f916afde14bbf436aa9a2196f2d4e278fb2454a", + "prepared_review_unit_hash": "0ac861be93c45f4c824fea12ed2e172a078d8d6d85014a2755c0245f53dc316f", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 25, - "serialized_chars": 34559, - "estimated_input_tokens": 8640, - "estimated_eval_tokens": 17536 + "soft_pair_count": 26, + "serialized_chars": 34793, + "estimated_input_tokens": 8699, + "estimated_eval_tokens": 17654 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is largely heterogeneous: most items are distinct model, tokenizer, loading, or training bugs with only superficial keyword overlap. No soft pair looks like the same underlying bug/change, so all candidate duplicate links should be rejected.", - "confidence": 0.22, - "canonical_issue_reason": "issue:43873 is the broadest and most referenced report in the set, so it is the best available cluster anchor even though the cluster itself is noisy.", + "summary": "This looks like a loose similarity cluster of mostly unrelated Transformers issues rather than true duplicates. The soft pairs share broad subsystem language (tokenizers, cache, generation, multimodal loading, quantization), but the concrete failure modes and code paths differ, so none should be merged as duplicates.", + "confidence": 0.21, + "canonical_issue_reason": "issue:43873 has the strongest triage signal in this set: it is active, has the most inbound references and discussion, and describes a concrete reproducible bug with broad user impact.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the most central and general issue here (open, multiple inbound references, broad quantization/offloading scope).", + "best_issue_reason": "issue:43873 is the most representative issue to anchor the cluster because it is well-discussed, externally referenced, and clearly scoped as a real bug report.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43441", - "right": "issue:47337", + "left": "issue:43716", + "right": "issue:46821", "accept": false, - "reason": "Different problem domains: FlashAttention/model loading on one side, flaky bbox-range test failure on the other." + "reason": "Different products and failures: image preprocessor dtype mismatch vs a window_size AttributeError. Not the same underlying bug." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Unrelated bugs: early-stopping metric lookup vs image processor handling of void segmentation maps." + "reason": "Early stopping/eval metric handling and void segmentation maps are unrelated code paths." }, { - "left": "issue:44186", - "right": "issue:44361", + "left": "issue:43441", + "right": "issue:47337", "accept": false, - "reason": "Both are tokenizer issues, but they concern different tokenizers and different failure modes; not the same bug." + "reason": "FlashAttention failure for a specific model/version is unrelated to a flaky bbox range test failure." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "Both are load failures, but for different models and different root causes; no evidence of a shared code-path bug." + "reason": "Both are loading issues, but they concern different models and likely different loader bugs; too broad to merge." }, { - "left": "issue:44734", - "right": "issue:46858", + "left": "issue:44186", + "right": "issue:44361", "accept": false, - "reason": "Both involve cache/continuation arithmetic, but the concrete failures differ and do not look mergeable into one fix." + "reason": "Tokenizer crashes, but on different tokenizer classes and different trigger conditions; not the same bug." }, { "left": "issue:44464", "right": "issue:46858", "accept": false, - "reason": "Compiled generation inconsistency is broader and different from the specific static-cache seq-length tensor bug." + "reason": "Both involve generation/cache/torch.compile, but one is inconsistent chunked generation output and the other is a static-cache length tensor arithmetic bug on non-CUDA backends." }, { - "left": "issue:30064", - "right": "issue:34634", + "left": "issue:44734", + "right": "issue:46858", "accept": false, - "reason": "Image segmentation-map processing and Bark voice_preset handling are unrelated." + "reason": "Serving KV-cache continuation indexing bug is distinct from the device-tensor cache length arithmetic issue." }, { - "left": "issue:43452", - "right": "issue:47020", + "left": "issue:30064", + "right": "issue:34634", "accept": false, - "reason": "Generic gguf/from_pretrained breakage is not the same as a specific XLMRobertaTokenizer constructor mismatch." + "reason": "Image segmentation-map processing and BarkProcessor voice_preset behavior are unrelated." }, { - "left": "issue:44671", - "right": "issue:44743", + "left": "issue:29159", + "right": "issue:36032", "accept": false, - "reason": "Different models and different symptom classes: masked-LM prediction correctness vs recurrent state reset with cache." + "reason": "Generic slow/fast tokenizer inconsistency is not the same as T5 tokenizer class-name conflict on load." }, { - "left": "issue:29159", - "right": "issue:36032", + "left": "issue:43452", + "right": "issue:47020", "accept": false, - "reason": "General slow-vs-fast tokenizer inconsistency does not match the specific T5 special-token method conflict." + "reason": "GGUF/autoloading breakage and an XLMRoberta Unigram constructor mismatch are different tokenizer/model-init bugs." }, { "left": "issue:42915", "right": "issue:43322", "accept": false, - "reason": "Separate model-loading failures with different backends and symptoms; too far apart to be duplicates." + "reason": "Qwen3Moe FP8 config failure and Llava Next loading segfault are different model-loading paths and failures." }, { - "left": "issue:44387", - "right": "issue:45701", + "left": "issue:44671", + "right": "issue:44743", "accept": false, - "reason": "Quantization memory/OOM regression and tokenization changes are unrelated." + "reason": "Masked LM prediction regression in CamemBERT is unrelated to Qwen recurrent-state reset behavior." }, { - "left": "issue:45479", - "right": "issue:47317", + "left": "issue:44387", + "right": "issue:45701", "accept": false, - "reason": "Both mention training loss, but one is a known num_labels=1 loss-selection bug and the other lacks evidence of the same mechanism." + "reason": "Quantization memory/OOM regression is not the same as a tokenization-version-change issue." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantization/offloading behavior and an apex RMSNorm memory leak are different bugs." + "reason": "Quantization offloading behavior and an apex FusedRMSNorm memory leak are distinct bugs." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and missing text_config unwrapping in composite multimodal configs are different implementation problems." }, { "left": "issue:33357", "right": "issue:35707", "accept": false, - "reason": "CLIP MacOS bus error and generation with inputs_embeds/past_key_values are unrelated." + "reason": "MacOS bus error on a CLIP model and progressive generation with inputs_embeds/past_key_values are unrelated." }, { - "left": "issue:43531", - "right": "issue:45759", + "left": "issue:45479", + "right": "issue:47317", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is a different code path from config unwrapping for composite Qwen multimodal models." + "reason": "Both are training-loss bugs, but one is a general single-label classification edge case and the other is PPFormulaNet-specific." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Different Qwen multimodal failures: video-input processing vs weight_only loading." + "reason": "Video input processing in Qwen3OmniMoe and Qwen3VL weight_only loading errors are different multimodal bugs." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Config API discrepancy for Gemma3TextConfig is not the same bug as GPTNeoX rotary_pct reload behavior." + "reason": "API discrepancy in Gemma3TextConfig is unrelated to GPTNeoX rotary_pct reload persistence." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "torch.compile+SDPA failure in Bamba is unrelated to BigBirdTokenizer special-token registration." - }, - { - "left": "issue:41720", - "right": "issue:43746", - "accept": false, - "reason": "A800 device-mapping CUDA assert and PEFT checkpoint loading are different issues." + "reason": "torch.compile+SDPA failure in Bamba and BigBirdTokenizer special-token registration are unrelated." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "Both concern file/path handling, but one is a cache-path collision and the other is permissions propagation on save." + "reason": "Cache-path collisions for trust_remote_code and read-only permission propagation on save_pretrained are different filesystem issues." }, { "left": "issue:44534", "right": "issue:46665", "accept": false, - "reason": "Non-persistent buffer initialization junk and from_pretrained behavior inside Python classes are not the same bug." + "reason": "Non-persistent buffer initialization junk and from_pretrained-in-Python-class behavior are not the same code-path problem." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3 device-mapping cudaErrorAssert and GraniteSpeech PEFT checkpoint loading are unrelated." }, { "left": "issue:46894", "right": "issue:47317", "accept": false, - "reason": "Both mention training loss, but 46894 is a specific double-shift ASR bug and 47317 has no evidence of that same pattern." + "reason": "Both mention training loss, but the models and underlying loss computation bugs differ; same class of symptom is not enough." }, { "left": "issue:43257", "right": "issue:47254", "accept": false, - "reason": "DeepSpeed/accelerate MOE weight conversion and PEFT ZeRO-3 checkpoint errors are separate failure modes." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and a PEFT/ZeRO-3 checkpoint error are different distributed-loading issues." }, { - "left": "issue:43606", - "right": "issue:43720", + "left": "issue:34634", + "right": "issue:41720", "accept": false, - "reason": "CPU offload device mismatch and BitNet packed-weight unpacking are different accelerate-loading problems." + "reason": "BarkProcessor voice_preset and Qwen3 auto device-mapping cuda assert are unrelated." } ] }, @@ -5577,11 +5673,11 @@ "nodes": [ "issue:29159", "issue:29925", + "issue:30064", "issue:34567", - "issue:34634", "issue:35707", + "issue:36010", "issue:39692", - "issue:41720", "issue:42490", "issue:43329", "issue:43388", @@ -5591,7 +5687,7 @@ "issue:43606", "issue:43644", "issue:43653", - "issue:43746", + "issue:43720", "issue:43881", "issue:43883", "issue:44062", @@ -5615,7 +5711,6 @@ "issue:45759", "issue:45859", "issue:45923", - "issue:46050", "issue:46402", "issue:46665", "issue:46821", @@ -5626,216 +5721,216 @@ ], "soft_pairs": [ "issue:45406|issue:45464", - "issue:34634|issue:41720", + "issue:43606|issue:43720", "issue:43531|issue:46821", "issue:42490|issue:43329", - "issue:29159|issue:29925", "issue:43531|issue:44589", + "issue:29159|issue:29925", "issue:34567|issue:35707", - "issue:44492|issue:44509", "issue:43881|issue:44991", - "issue:43606|issue:43653", + "issue:44492|issue:44509", "issue:45759|issue:45859", + "issue:43606|issue:43653", "issue:43540|issue:46858", "issue:44451|issue:44991", "issue:45704|issue:47405", "issue:43388|issue:45059", "issue:43526|issue:43883", "issue:44987|issue:45042", - "issue:45923|issue:47333", "issue:45042|issue:47448", + "issue:45923|issue:47333", "issue:44077|issue:44625", "issue:43644|issue:46665", - "issue:45357|issue:46402", "issue:44898|issue:45072", + "issue:45357|issue:46402", "issue:39692|issue:44610", "issue:44062|issue:44451", - "issue:43746|issue:46050" + "issue:30064|issue:36010" ], - "prepared_review_unit_hash": "e8d2ea87530556ff791ad703c85b44cef83c9d946c0d42ba62e317adc40016d6", + "prepared_review_unit_hash": "07c20916d79d5b13e0f28a60fc71cf94a741eecf893ff0872ece51aff99fa754", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35626, - "estimated_input_tokens": 8907, - "estimated_eval_tokens": 18070 + "serialized_chars": 35071, + "estimated_input_tokens": 8768, + "estimated_eval_tokens": 17792 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All soft pairs are false positives: they share surface wording or a broad subsystem, but not the same concrete bug/change. No PRs were present. The set is heterogeneous, so there is no safe global duplicate canonical.", - "confidence": 0.92, - "canonical_issue_reason": null, + "summary": "This cluster is a mixed bag of unrelated Transformers issues; every soft pair looks like thematic similarity rather than the same underlying bug/change, so no duplicates are accepted.", + "confidence": 0.93, + "canonical_issue_reason": "No single canonical issue emerges; the items span unrelated tokenizer, loading, multimodal, training, and docs problems.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45406 is the clearest, most actionable regression report in the set: it names a specific failing code path (`Gemma4Processor` missing `_tokenizer`) and is narrowly scoped. It is a good triage anchor, but not a duplicate hub for the rest of the items.", + "best_issue_reason": "No global best issue; the set is too heterogeneous to represent one bug/change.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45406", "right": "issue:45464", "accept": false, - "reason": "Different bugs and code paths: Gemma4Processor `_tokenizer` crash in serve vs Qwen3.5 streaming chat/completions failure." + "reason": "Gemma4Processor missing `_tokenizer` in serve vs Qwen3.5 streaming chat/completions; different failure paths and components." }, { - "left": "issue:34634", - "right": "issue:41720", + "left": "issue:43606", + "right": "issue:43720", "accept": false, - "reason": "Bark `voice_preset` handling vs Qwen3 auto device mapping CUDA assert; same general area only at the broadest level." + "reason": "Both involve loading/offload, but one is a CPU offload device mismatch and the other is BitNet packed-weight unpacking during accelerate loading." }, { "left": "issue:43531", "right": "issue:46821", "accept": false, - "reason": "Both mention window/sliding-window terms, but one is Qwen3-MoE config behavior and the other is TimesFM `window_size` AttributeError; different models and fixes." + "reason": "Qwen3-MoE sliding_window behavior and TimesFM `window_size` AttributeError are unrelated model-specific bugs." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "Trailing-slash local path import warning is unrelated to a multimodal video token counting bug." - }, - { - "left": "issue:29159", - "right": "issue:29925", - "accept": false, - "reason": "Tokenizer inconsistency in general vs SigLIP specifically not enforcing `use_fast=True`; related theme, not the same bug." + "reason": "Trailing-slash local model import warning is unrelated to a multimodal video-token counting bug." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Qwen3-MoE sliding-window issue is unrelated to the float8 storage loading error." + "reason": "Sliding window handling in Qwen3-MoE is unrelated to the Float8 storage lookup error." }, { - "left": "issue:34567", - "right": "issue:35707", + "left": "issue:29159", + "right": "issue:29925", "accept": false, - "reason": "Trainer token counter bug vs generation with `inputs_embeds`/`past_key_values`; different trainer/generation paths." + "reason": "General slow/fast tokenizer inconsistency is not the same as SigLIP-specific enforcement of `use_fast=True`." }, { - "left": "issue:44492", - "right": "issue:44509", + "left": "issue:34567", + "right": "issue:35707", "accept": false, - "reason": "A typo in cache strategy docs vs deprecated pipeline-task documentation; both docs-ish but unrelated." + "reason": "Trainer token-count tracking and progressive generation with `inputs_embeds`/`past_key_values` are different code paths." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "GLM-4V loading failure vs tokenizer loading failure for EMBEDDIA/est-roberta; different models and failure points." + "reason": "GLM-4V model loading failure and est-roberta tokenizer loading failure are distinct model-specific issues." }, { - "left": "issue:43606", - "right": "issue:43653", + "left": "issue:44492", + "right": "issue:44509", "accept": false, - "reason": "CPU offload device mismatch in Bark vs BigBird tokenizer special-token registration; unrelated." + "reason": "A cache-strategy typo and stale docs about removed pipelines are unrelated documentation issues." }, { "left": "issue:45759", "right": "issue:45859", "accept": false, - "reason": "Composite Qwen config unwrap bug vs CPU-offload incompatibility in rotary embedding; different code paths." + "reason": "Unwrapping `text_config` for composite Qwen multimodal configs is unrelated to Qwen3_5Moe rotary embedding CPU-offload incompatibility." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to BigBirdTokenizer mask token special-token registration." }, { "left": "issue:43540", "right": "issue:46858", "accept": false, - "reason": "Video-input validation error in Qwen3OmniMoe vs static-cache slice arithmetic bug on non-CUDA backends." + "reason": "Qwen3OmniMoe video-input validation bug and static-cache seq-length tensor arithmetic bug are separate failures." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are model loading problems, but ScandiBERT loading and est-roberta tokenizer loading are distinct issues." + "reason": "ScandiBERT load failure and est-roberta tokenizer load failure are different checkpoints and symptoms." }, { "left": "issue:45704", "right": "issue:47405", "accept": false, - "reason": "Apex RMSNorm memory leak vs silent wrong-auto-class conversion skip; no shared bug mechanism." + "reason": "T5 apex RMSNorm memory leak is unrelated to silent model_type mismatch conversion skipping." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric gathering drops tuple labels vs SAM3 text/box behavior; unrelated training/eval vs model input bug." + "reason": "`gather_for_metrics` tuple-label truncation and SAM3 text/box odd behavior are different training/inference problems." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "Beit fast image-processor label reduction vs Molmo missing `all_tied_weights_keys`; completely different subsystems." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to Molmo missing `all_tied_weights_keys`." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Loading a specific converted model fails vs PIL backends incorrectly requiring torchvision; distinct causes and fixes." + "reason": "A model-loading failure for physical-intelligence/fast is not the same as PIL processors incorrectly requiring torchvision." }, { - "left": "issue:45923", - "right": "issue:47333", + "left": "issue:45042", + "right": "issue:47448", "accept": false, - "reason": "Missing gradient-checkpointing flag vs inability to save a disk-offloaded Qwen3 model; not the same problem." + "reason": "Torchvision dependency regression in PIL processors is unrelated to Gemma4 architecture recognition." }, { - "left": "issue:45042", - "right": "issue:47448", + "left": "issue:45923", + "right": "issue:47333", "accept": false, - "reason": "Processor dependency issue vs Gemma4 architecture recognition; only both are model/processor related." + "reason": "Missing `supports_gradient_checkpointing` and disk-offload save failure are different model lifecycle bugs." }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "PatchTSMixer `post_init` restriction vs Qwen3.5 `num_labels` propagation; unrelated model config bugs." + "reason": "patchtsmixer post-init policy and Qwen3.5 label propagation concern unrelated configs." }, { "left": "issue:43644", "right": "issue:46665", "accept": false, - "reason": "Non-persistent buffer initialization bug vs inconsistent `from_pretrained` behavior inside Python classes; both involve loading, but not the same defect." + "reason": "Non-persistent buffer initialization junk and `from_pretrained` inside Python classes are unrelated behaviors." }, { - "left": "issue:45357", - "right": "issue:46402", + "left": "issue:44898", + "right": "issue:45072", "accept": false, - "reason": "Incorrect visual encoder keys on save vs text encoder weights not loaded; different save/load directions and components." + "reason": "Perceiver non-default-resolution failure and bfloat16 dtype mismatches are separate model/backend issues." }, { - "left": "issue:44898", - "right": "issue:45072", + "left": "issue:45357", + "right": "issue:46402", "accept": false, - "reason": "Perceiver resolution/interpolation failure vs bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "Qwen3.5 `save_pretrained` key regression is unrelated to Sam3Model text encoder weights not loading." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "SigLIP2 doc/example issues vs OmDet-Turbo processor output-size mismatch; one is docs/example, the other is a processor-model contract bug." + "reason": "SigLIP2 documentation example errors are unrelated to OmDet-Turbo processor/model resolution mismatch." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "`AddedToken` keyword conflict vs model loading failure for ScandiBERT; unrelated tokenizer constructor vs checkpoint loading bugs." + "reason": "AddedToken keyword duplication and ScandiBERT loading failure are unrelated tokenizer/model issues." }, { - "left": "issue:43746", - "right": "issue:46050", + "left": "issue:30064", + "right": "issue:36010", "accept": false, - "reason": "PEFT local-checkpoint loading bug in GraniteSpeech vs quantized-model save/load corruption; different persistence paths." + "reason": "Void segmentation-map processing and `GenerationMixin` import error have no shared underlying code path." } ] }, @@ -5859,14 +5954,12 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ - "issue:30064", - "issue:36010", "issue:39692", "issue:42886", "issue:43381", "issue:43421", - "issue:43493", - "issue:43653", + "issue:43540", + "issue:43596", "issue:43673", "issue:43720", "issue:43746", @@ -5879,7 +5972,7 @@ "issue:43931", "issue:43994", "issue:44077", - "issue:44186", + "issue:44220", "issue:44368", "issue:44464", "issue:44514", @@ -5898,6 +5991,8 @@ "issue:45479", "issue:45907", "issue:46032", + "issue:46050", + "issue:46294", "issue:46463", "issue:46655", "issue:46693", @@ -5908,237 +6003,230 @@ "issue:47458" ], "soft_pairs": [ - "issue:30064|issue:36010", "issue:44918|issue:44936", + "issue:44514|issue:44545", + "issue:43746|issue:46050", "issue:42886|issue:44843", - "issue:43746|issue:43873", - "issue:44077|issue:47458", "issue:44625|issue:45479", - "issue:44514|issue:44545", - "issue:43421|issue:43927", + "issue:43746|issue:43873", "issue:46655|issue:46693", - "issue:44995|issue:46858", "issue:44464|issue:47050", + "issue:44995|issue:46858", + "issue:44077|issue:47458", + "issue:39692|issue:43873", "issue:43381|issue:43844", - "issue:39692|issue:43493", - "issue:43931|issue:47458", - "issue:43673|issue:47337", - "issue:43782|issue:45381", "issue:45081|issue:45356", - "issue:43421|issue:43653", - "issue:44368|issue:45357", + "issue:43782|issue:45381", + "issue:43673|issue:47337", "issue:44857|issue:47050", "issue:46032|issue:46463", + "issue:43931|issue:47458", "issue:43881|issue:47448", + "issue:44368|issue:45357", "issue:43994|issue:44671", - "issue:43421|issue:44186", - "issue:39692|issue:43873", "issue:39692|issue:47050", + "issue:43421|issue:43927", "issue:45356|issue:45907", - "issue:43720|issue:43819" + "issue:43540|issue:46294", + "issue:43720|issue:43819", + "issue:43596|issue:44220" ], - "prepared_review_unit_hash": "b6b2100dc55d7e1648aa69023c7ee761d2c5a5c216fe169cbf2f67a6d698f951", + "prepared_review_unit_hash": "a39f79417584fcbea8573271720cf1b1ef4b07b3725b44f543fb1204059f305c", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 28, - "serialized_chars": 35848, - "estimated_input_tokens": 8962, - "estimated_eval_tokens": 18180 + "soft_pair_count": 27, + "serialized_chars": 35543, + "estimated_input_tokens": 8886, + "estimated_eval_tokens": 18028 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are unrelated issue reports across different models/subsystems. The only clear duplicate soft edge is the Qwen2_5_VLProcessor.apply_chat_template crash on batched input with padding=False; the later report is a duplicate of the earlier one. No PRs are present.", - "confidence": 0.86, - "canonical_issue_reason": "issue:44514 is the earlier, cleaner report of the Qwen2_5_VLProcessor.apply_chat_template batched-input crash with padding=False, so it is the better canonical issue.", + "summary": "This cluster is mostly a set of unrelated bug reports that only weakly resemble each other by broad themes like tokenizers, model loading, caching, or multimodal preprocessing. Only two soft pairs look like true duplicates or near-duplicates: the Qwen2.5-VL batched padding crash pair, and the HF_HUB_OFFLINE tokenizer regression pair.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue cleanly covers the whole cluster, but issue 44843 is the best anchor because it identifies a concrete offline-tokenizer regression and its likely root cause in _patch_mistral_regex/model_info. The exact-title duplicate pair 44514/44545 is narrower and less representative of the cluster as a whole.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44514 is the strongest representative issue because it precisely identifies the failing code path and has a near-identical later duplicate.", + "best_issue_reason": "Issue 44843 is the strongest standalone candidate: it is specific, reproducible, and points to a clear code path, and it aligns with one of the accepted soft duplicates. Issue 44545 is an exact duplicate of 44514, but it is narrower in scope.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:30064", - "right": "issue:36010", - "accept": false, - "reason": "Different failures: segmentation-map handling in Image Processor vs GenerationMixin import error." - }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Different bugs: Qwen3.5 embedding unpacking with TRL SFT trainer vs trainer.evaluate() failing after train()." + "reason": "Both involve training/evaluation behavior, but they describe different failures and likely different code paths." }, { - "left": "issue:42886", - "right": "issue:44843", - "accept": false, - "reason": "Both are offline-related, but they hit different code paths: tokenizer cache loading vs _patch_mistral_regex calling model_info()." + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same bug: identical Qwen2_5_VLProcessor.apply_chat_template crash on batched input with padding=False." }, { "left": "issue:43746", - "right": "issue:43873", + "right": "issue:46050", "accept": false, - "reason": "Different issues: PEFT/local checkpoint loading for GraniteSpeech vs quantization/offloading behavior." + "reason": "Both are save/load-related, but one is PEFT local checkpoint loading and the other is post-save quantized model corruption." }, { - "left": "issue:44077", - "right": "issue:47458", - "accept": false, - "reason": "Unrelated model internals: patchtsmixer post_init policy vs GPT2 c_proj initialization." + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same underlying offline tokenizer regression; 44843 pins the root cause to _patch_mistral_regex/model_info." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "Different problems: num_labels propagation in Qwen3.5 vs degenerate zero loss in single-label classification." + "reason": "Both concern classification configs, but one is num_labels propagation and the other is a degenerate loss formulation." }, { - "left": "issue:44514", - "right": "issue:44545", - "accept": true, - "reason": "Same underlying crash in Qwen2_5_VLProcessor.apply_chat_template on batched input with padding=False; wording differs but the code-path failure is identical." - }, - { - "left": "issue:43421", - "right": "issue:43927", + "left": "issue:43746", + "right": "issue:43873", "accept": false, - "reason": "Different tokenizer state bugs: runtime special-token post-processor updates vs DiaConfig losing custom token IDs on save/load." + "reason": "Local checkpoint loading with adapters is not the same as quantization/offloading behavior." }, { "left": "issue:46655", "right": "issue:46693", "accept": false, - "reason": "Both mention generation/performance, but one is a video-processor padding bug and the other is an FA2 long-context performance regression." - }, - { - "left": "issue:44995", - "right": "issue:46858", - "accept": false, - "reason": "Different cache issues: stale indexer cache on second forward pass vs device-tensor seq_length breaking static-cache compile arithmetic." + "reason": "Video padding calculation and FA2 long-context performance regression are different problems." }, { "left": "issue:44464", "right": "issue:47050", "accept": false, - "reason": "Both involve incorrect outputs, but one is chunked generation with compiled forward and the other is ONNX fp16 attention dtype mismatch." - }, - { - "left": "issue:43381", - "right": "issue:43844", - "accept": false, - "reason": "Different training/runtime bugs: gradient checkpointing in eval mode vs abnormal gradients with HfDeepSpeedConfig + ZeRO-3." + "reason": "Chunked generation inconsistency under compiled forward is unrelated to ONNX fp16 attention dtype mismatch." }, { - "left": "issue:39692", - "right": "issue:43493", + "left": "issue:44995", + "right": "issue:46858", "accept": false, - "reason": "Same model family, but one is a documentation example mismatch/quantization issue and the other is an implementation discrepancy vs JAX." + "reason": "Both mention cache, but one is a stale indexer cache crash and the other is a static-cache tensor-type arithmetic issue." }, { - "left": "issue:43931", + "left": "issue:44077", "right": "issue:47458", "accept": false, - "reason": "Different loading/shape problems: Qwen3-VL weight-shape mismatch vs GPT2 c_proj initialization." + "reason": "Post-init allowance in patchtsmixer is unrelated to GPT2 c_proj initialization." }, { - "left": "issue:43673", - "right": "issue:47337", + "left": "issue:39692", + "right": "issue:43873", "accept": false, - "reason": "Different bugs: missing GenerationMixin cache in chunked_prefill vs a flaky bbox-range test error." + "reason": "Documentation/example errors are unrelated to quantization/offloading behavior." }, { - "left": "issue:43782", - "right": "issue:45381", + "left": "issue:43381", + "right": "issue:43844", "accept": false, - "reason": "Both are Qwen/VL-related, but one is a weight_only load error and the other is wrong vision_position_ids for video input." + "reason": "Gradient checkpointing in eval mode and gradient growth under ZeRO-3 are different issues." }, { "left": "issue:45081", "right": "issue:45356", "accept": false, - "reason": "Related tokenizer/regression area, but different concrete failures: backend_tokenizer AttributeError vs Kimi-K2.5 codec handling regression." + "reason": "Both touch tokenizer regex handling, but one is a crash and the other is a regression around codec handling/warnings." }, { - "left": "issue:43421", - "right": "issue:43653", + "left": "issue:43782", + "right": "issue:45381", "accept": false, - "reason": "Both concern tokenizer special tokens, but one is runtime post-processor updates and the other is BigBird mask-token registration/empty decode output." + "reason": "Different Qwen/VL problems with different symptoms and code paths." }, { - "left": "issue:44368", - "right": "issue:45357", + "left": "issue:43673", + "right": "issue:47337", "accept": false, - "reason": "Different concerns: a warning about tie_word_embeddings in LoRA fine-tuning vs incorrect visual-encoder keys saved in Qwen3.5." + "reason": "Generation cache behavior and a bbox-range flaky test are unrelated." }, { "left": "issue:44857", "right": "issue:47050", "accept": false, - "reason": "Different tensor/precision bugs: LwDetrImageLoss AMP crash vs CanineModel fp16 ONNX attention dtype mismatch." + "reason": "Different models and different failure modes; only a superficial fp16/CUDA theme is shared." }, { "left": "issue:46032", "right": "issue:46463", "accept": false, - "reason": "Different code paths: Mamba2Mixer use_cache correctness vs T5 decoder-input validation error messaging." + "reason": "Cache correctness in Mamba2Mixer is unrelated to a T5 error-message bug." + }, + { + "left": "issue:43931", + "right": "issue:47458", + "accept": false, + "reason": "Model shape mismatch during load is unrelated to GPT2 weight initialization." }, { "left": "issue:43881", "right": "issue:47448", "accept": false, - "reason": "Different model-loading issues: glm-4v-9b loading failure vs Gemma4 architecture recognition." + "reason": "Both are model-loading problems, but with different models and different causes." }, { - "left": "issue:43994", - "right": "issue:44671", + "left": "issue:44368", + "right": "issue:45357", "accept": false, - "reason": "Both involve model behavior regressions, but one is SigLIP2 inference inconsistency and the other is CamemBERT masked-LM prediction errors." + "reason": "A warning about tie_word_embeddings is unrelated to incorrect visual encoder keys on save_pretrained." }, { - "left": "issue:43421", - "right": "issue:44186", + "left": "issue:43994", + "right": "issue:44671", "accept": false, - "reason": "Different tokenizer bugs: runtime special-token post-processor updates vs LayoutLMv2Tokenizer crashes on NER/batched padding." + "reason": "Both affect model outputs, but SigLIP2 pipeline issues and CamemBERT MLM regression are distinct." }, { "left": "issue:39692", - "right": "issue:43873", + "right": "issue:47050", "accept": false, - "reason": "Same broad SigLIP2/quantization area is not enough; these are different problems, with one about example errors and the other about offloading with quantization." + "reason": "Documentation example problems do not match an ONNX export dtype mismatch." }, { - "left": "issue:39692", - "right": "issue:47050", + "left": "issue:43421", + "right": "issue:43927", "accept": false, - "reason": "Different failures: SigLIP2 example/model discrepancy vs fp16 ONNX export dtype mismatch." + "reason": "Runtime post-processor updates and DiaConfig token-ID loss are separate tokenizer/config bugs." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Different regressions: tokenizer codec/fix_mistral_regex issue vs list_repo_files not forwarding revision." + "reason": "Tokenizer codec/regression handling is unrelated to list_repo_files revision forwarding." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-input failures, but they involve different models and different failure mechanisms." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Different BitNet/DAC problems: packed-weight loading during accelerate vs DAC from_latents missing STE mismatch." + "reason": "BitNet packed-weight loading and DAC latent conversion/STE mismatch are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "DeepSpeed/BertModel zero3 init crash is unrelated to fbank feature extraction." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: all cited items are issues, and the only clear duplicate pair is issue:44514 / issue:44545 with matching titles and closely aligned descriptions. The other soft-edge verdicts are stated conservatively as non-duplicates based on different titles/code paths. No PR pair judgment is needed because the packet contains no PRs." + "feedback": "The cluster summary is grounded in the packet and stays conservative. The two accepted soft pairs are well-supported by the issue titles and descriptions, and the rejected pairs are framed cautiously rather than overclaimed. No obvious evidence inflation stands out." }, "evaluator_used": true, "retried": false, @@ -6163,13 +6251,14 @@ "issue:34634", "issue:36010", "issue:36331", + "issue:39692", "issue:39921", "issue:41720", "issue:43295", + "issue:43421", "issue:43531", "issue:43540", "issue:43550", - "issue:43596", "issue:43611", "issue:43643", "issue:43653", @@ -6196,7 +6285,6 @@ "issue:45381", "issue:45834", "issue:45987", - "issue:46294", "issue:46650", "issue:46665", "issue:46693", @@ -6209,248 +6297,252 @@ "issue:47458" ], "soft_pairs": [ - "issue:43540|issue:46294", + "issue:44821|issue:45084", "issue:43550|issue:43819", - "issue:43596|issue:44220", "issue:43550|issue:43927", + "issue:43421|issue:44186", + "issue:43421|issue:43653", "issue:43295|issue:45834", - "issue:44821|issue:45084", - "issue:44488|issue:47020", "issue:45072|issue:47050", "issue:30064|issue:36331", - "issue:44062|issue:44488", + "issue:43540|issue:44008", + "issue:44488|issue:47020", "issue:43653|issue:43720", + "issue:44062|issue:44488", "issue:45381|issue:46693", "issue:41720|issue:44991", + "issue:39921|issue:44936", "issue:43531|issue:43643", - "issue:43540|issue:44008", "issue:44220|issue:44661", - "issue:39921|issue:44936", "issue:43611|issue:46665", "issue:46829|issue:46832", - "issue:44863|issue:47458", - "issue:43653|issue:43819", - "issue:44186|issue:44898", + "issue:44008|issue:46858", "issue:34634|issue:36010", + "issue:44186|issue:44898", + "issue:43653|issue:43819", + "issue:44863|issue:47458", "issue:46650|issue:47405", "issue:44991|issue:47020", - "issue:44008|issue:46858", "issue:44843|issue:45081", "issue:44561|issue:45020", + "issue:39692|issue:45834", "issue:45987|issue:46829" ], - "prepared_review_unit_hash": "1d1fcd138530588b1f70bc34a79ae67fa3ea7800367ba491028b32bd97315867", + "prepared_review_unit_hash": "2da3904f7c13b22b4da22b60e221c3d62b9eb692160bae1439b15393936fac17", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36769, - "estimated_input_tokens": 9193, - "estimated_eval_tokens": 18642 + "soft_pair_count": 30, + "serialized_chars": 37063, + "estimated_input_tokens": 9266, + "estimated_eval_tokens": 18788 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most candidates are clearly distinct bugs across different models, APIs, or runtime paths. The only plausible duplicate pair is the two Mistral `_patch_mistral_regex` reports, which hit the same helper during tokenizer loading and could likely be fixed together.", - "confidence": 0.83, - "canonical_issue_reason": "Issue 44843 is the best canonical issue: it targets the shared `_patch_mistral_regex` code path at the broadest/root-cause level, while 45081 looks like a narrower symptom in the same area.", + "summary": "The set is highly heterogeneous: the soft pairs are mostly same-subsystem but not same-bug matches, so this cluster does not support duplicate merging. I found one near-adjacent tokenizer-regex pair, but even that looks like two distinct failures in the same helper rather than one duplicate.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated bugs across tokenizers, vision, training, model loading, and cache behavior, so there is no single duplicate family to anchor on.", "canonical_pr_reason": null, - "best_issue_reason": "44843 is the most actionable and general report in the only credible duplicate pair, with a clear failure mode and a likely shared fix surface.", + "best_issue_reason": "No issue is a strong global representative because the approximate matches are mostly false positives; the closest items still describe different concrete failure modes.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43540", - "right": "issue:46294", + "left": "issue:44821", + "right": "issue:45084", "accept": false, - "reason": "Both mention video input, but they are different Qwen variants and different failure modes (`ValueError` vs chat-template feature/token mismatch)." + "reason": "Both are image/tokenizer-related only at a very broad level; one is AutoImageProcessor URL loading, the other is a compile/template-node error. Different code paths and fixes." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Unrelated bugs: torch.compile/SDPA in Bamba versus DAC latent reconstruction/STE behavior." - }, - { - "left": "issue:43596", - "right": "issue:44220", - "accept": false, - "reason": "Different subsystems and symptoms: Deepspeed Zero3/BertModel indexing versus audio fbank feature extraction." + "reason": "Different model families and failures: SDPA/torch.compile on Bamba vs missing STE in DAC.from_latents. Not the same underlying bug." }, { "left": "issue:43550", "right": "issue:43927", "accept": false, - "reason": "Different model/runtime issues: Bamba compile failure versus DiaConfig token ID persistence/generation error." + "reason": "One is a compile/SDPA failure, the other is token ID loss after save/load in DiaConfig. Unrelated root causes." }, { - "left": "issue:43295", - "right": "issue:45834", + "left": "issue:43421", + "right": "issue:44186", "accept": false, - "reason": "Both are multimodal, but one is a processor regression and the other is a Kosmos OCR long-input indexing bug." + "reason": "Both involve tokenizers, but one is runtime post-processor updates and the other is LayoutLMv2 NER/padding crashes. Different behavior and likely different fixes." }, { - "left": "issue:44821", - "right": "issue:45084", + "left": "issue:43421", + "right": "issue:43653", "accept": false, - "reason": "No shared concrete bug: loading an AutoImageProcessor from URL versus a template-compilation type error." + "reason": "Post-processor update bug vs BigBird mask token registration/empty decode. Same area, but not the same concrete defect." }, { - "left": "issue:44488", - "right": "issue:47020", + "left": "issue:43295", + "right": "issue:45834", "accept": false, - "reason": "Both are tokenizer-loading related, but the affected models and failure mechanisms differ substantially." + "reason": "Processor.tokenizer regression vs Kosmos2.5 long-OCR index error. Different models and failure mechanisms." }, { "left": "issue:45072", "right": "issue:47050", "accept": false, - "reason": "Both mention dtype mismatches, but one is bfloat16 inference in SwitchTransformers/TimmWrapperModel and the other is fp16 ONNX export in CanineModel." + "reason": "Both mention dtype mismatches, but in different models and pipelines (SwitchTransformers/TimmWrapperModel inference vs CanineModel ONNX export). Not mergeable as one fix." }, { "left": "issue:30064", "right": "issue:36331", "accept": false, - "reason": "Completely different areas: image segmentation processing versus custom trainer loss signature compatibility." + "reason": "Void segmentation map processing vs CustomTrainer.compute_loss signature break. Completely unrelated." }, { - "left": "issue:44062", - "right": "issue:44488", + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Video-input ValueError in Qwen3OmniMoe vs Gemma3n variable-name collision in forward(). Different models and bugs." + }, + { + "left": "issue:44488", + "right": "issue:47020", "accept": false, - "reason": "AddedToken keyword conflict is unrelated to tokenizer loading failure for a specific model." + "reason": "Both are tokenizer-loading problems, but one is a model-specific load failure and the other is an XLMRobertaTokenizer constructor type mismatch. Not the same issue." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "Different bug classes: tokenizer special-token registration versus BitNet packed-weight unpacking during accelerate loading." + "reason": "BigBird special-token decoding bug vs BitNet packed-weight unpacking during accelerate loading. Different subsystems and fixes." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken special-argument duplication vs loading a specific tokenizer model. No shared concrete failure." }, { "left": "issue:45381", "right": "issue:46693", "accept": false, - "reason": "Video position IDs in Qwen2.5-VL is unrelated to long-context FA2 performance regression." + "reason": "Qwen2.5-VL video position IDs vs FA2 long-context performance regression. Different symptoms and code paths." }, { "left": "issue:41720", "right": "issue:44991", "accept": false, - "reason": "Different problems: device-mapping cuda assert in Qwen3 versus tokenizer loading failure for EMBEDDIA/est-roberta." + "reason": "CUDA assert on Qwen3 auto device mapping vs tokenizer loading regression for est-roberta. Unrelated." }, { - "left": "issue:43531", - "right": "issue:43643", + "left": "issue:39921", + "right": "issue:44936", "accept": false, - "reason": "One is a Qwen3-MoE sliding-window issue; the other is a `trust_remote_code` config-field loss issue." + "reason": "Gemma3N special-token projection error vs trainer.evaluate after train. No meaningful overlap." }, { - "left": "issue:43540", - "right": "issue:44008", + "left": "issue:43531", + "right": "issue:43643", "accept": false, - "reason": "Different models and errors: video processing failure in Qwen3OmniMoe versus Gemma3n forward variable collision." + "reason": "Sliding-window bug in Qwen3-MoE vs trust_remote_code AutoConfig missing fields. Different functionality." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio fbank feature extraction and `add-new-model-like` tokenizer mapping behavior are unrelated." - }, - { - "left": "issue:39921", - "right": "issue:44936", - "accept": false, - "reason": "Gemma3N special-token projection error is unrelated to trainer evaluate/train state behavior." + "reason": "Audio fbank feature extraction issue vs add-new-model-like failure in TOKENIZER_MAPPING_NAMES. Unrelated." }, { "left": "issue:43611", "right": "issue:46665", "accept": false, - "reason": "Both concern loading, but one is a v5 `base_model_prefix` breakage and the other is a generic from_pretrained-in-class inconsistency." + "reason": "Both concern model loading, but one is a base_model_prefix regression and the other is inconsistent from_pretrained behavior inside Python classes. Not the same concrete bug." }, { "left": "issue:46829", "right": "issue:46832", "accept": false, - "reason": "Different model heads and failures: output field misassignment versus RT-DETR feature-level mismatch." + "reason": "VideoPrism output field mismatch vs RT-DETR forward TypeError from feature-level count. Different model heads and failures." }, { - "left": "issue:44863", - "right": "issue:47458", + "left": "issue:44008", + "right": "issue:46858", "accept": false, - "reason": "Nemotron checkpoint loading failure and GPT2 c_proj initialization are different model-specific bugs." + "reason": "Gemma3n audio_mel_mask name collision vs static-cache seq-length tensor arithmetic on non-CUDA backends. Unrelated." }, { - "left": "issue:43653", - "right": "issue:43819", + "left": "issue:34634", + "right": "issue:36010", "accept": false, - "reason": "Tokenizer special-token registration and DAC latent conversion mismatch are unrelated." + "reason": "BarkProcessor voice_preset vs GenerationMixin import breakage. Different APIs and root causes." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "LayoutLMv2 tokenizer NER/padding bug and Perceiver non-default resolution bug are different code paths." + "reason": "LayoutLMv2 tokenizer crash on NER/padding vs Perceiver non-default resolution failure. Only broad vision/tokenizer similarity." }, { - "left": "issue:34634", - "right": "issue:36010", + "left": "issue:43653", + "right": "issue:43819", "accept": false, - "reason": "BarkProcessor voice preset handling and GenerationMixin import failure are unrelated." + "reason": "BigBird special-token handling vs DAC from_latents/STE mismatch. Different model internals and fixes." + }, + { + "left": "issue:44863", + "right": "issue:47458", + "accept": false, + "reason": "NemotronH checkpoint loading failure vs GPT2 c_proj weight initialization bug. Distinct model families and failure modes." }, { "left": "issue:46650", "right": "issue:47405", "accept": false, - "reason": "One is a non-idempotent save_pretrained rename bug; the other is a wrong-Auto-class silent conversion/load issue." + "reason": "Non-idempotent save_pretrained rename bug vs silent model_type mismatch conversion skip. Both serialization-related, but not the same defect." }, { "left": "issue:44991", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer load failures, but they involve different tokenizer implementations and distinct root causes." - }, - { - "left": "issue:44008", - "right": "issue:46858", - "accept": false, - "reason": "Gemma3n forward naming collision is unrelated to static cache sequence-length tensor behavior." + "reason": "Both are tokenizer loading issues, but for different models and with different failure points. Not a duplicate pair." }, { "left": "issue:44843", "right": "issue:45081", - "accept": true, - "reason": "Both hit the same `_patch_mistral_regex` path during Mistral tokenizer loading, and a single fix to that helper could plausibly address both the offline model_info call and the backend_tokenizer attribute crash." + "accept": false, + "reason": "Both hit _patch_mistral_regex, but one is an offline-mode network call bug and the other is an AttributeError on tokenizer shape. Same helper, different concrete bugs." }, { "left": "issue:44561", "right": "issue:45020", "accept": false, - "reason": "Both mention remote-code breakage, but they point to different regressions and different missing functionality." + "reason": "Both mention trust_remote_code / v5 loading regressions, but one is removal of is_torch_fx_available and the other is broader remote-code breakage. Too broad to merge." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "SigLIP2 doc/example errors vs Kosmos2.5 OCR index error. Different scopes and fixes." }, { "left": "issue:45987", "right": "issue:46829", "accept": false, - "reason": "StaticCache sequence-length return type and video classification output fields are unrelated." + "reason": "StaticCache.get_seq_length return-type bug vs VideoPrism hidden_states output mismatch. Not the same code path." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The assessment is conservative and grounded in the packet. The only accepted pair is well-supported by both issues pointing to `_patch_mistral_regex` in the Mistral tokenizer-loading path, and the remaining soft-edge rejections stay cautious about differing models, subsystems, or failure modes. The summary does not overclaim beyond the 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 }, @@ -6467,7 +6559,6 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ - "issue:39692", "issue:43322", "issue:43335", "issue:43421", @@ -6517,217 +6608,210 @@ "issue:47458" ], "soft_pairs": [ - "issue:39692|issue:45834", "issue:43643|issue:44877", "issue:43335|issue:43756", - "issue:45987|issue:46932", "issue:44448|issue:44671", + "issue:45987|issue:46932", "issue:44186|issue:44442", + "issue:43716|issue:44008", "issue:43322|issue:45685", "issue:44991|issue:46291", - "issue:43618|issue:43697", - "issue:43716|issue:44008", "issue:46294|issue:46463", "issue:44805|issue:44936", + "issue:43618|issue:43697", "issue:43606|issue:45072", "issue:45584|issue:45834", "issue:43550|issue:45072", "issue:44479|issue:45834", + "issue:43782|issue:44977", "issue:45183|issue:45479", "issue:46752|issue:47337", - "issue:43782|issue:44977", - "issue:44534|issue:47458", - "issue:43421|issue:44811", - "issue:43577|issue:43818", "issue:43644|issue:43749", "issue:45563|issue:47276", "issue:43582|issue:46512", - "issue:44610|issue:45072" + "issue:43577|issue:43818", + "issue:43421|issue:44811", + "issue:44610|issue:45072", + "issue:44534|issue:47458" ], - "prepared_review_unit_hash": "734aac357787418ef124364e6edf21947dd7e4ba2ecb76eab051a5696e4080a2", + "prepared_review_unit_hash": "7cbac71cea774470ef8835aacdcff31971adbd1fb5a9ef69752915106b64a0b6", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 26, - "serialized_chars": 35658, - "estimated_input_tokens": 8915, - "estimated_eval_tokens": 18086 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 25, + "serialized_chars": 34764, + "estimated_input_tokens": 8691, + "estimated_eval_tokens": 17638 }, "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, tokenizer, config, dtype, or generation regressions. None of the soft edges look like the same underlying bug/change strongly enough to merge, so all candidate links are rejected. The only recurring theme is broad Transformers v5 regressions, but the concrete failure modes still differ.", - "confidence": 0.94, - "canonical_issue_reason": "No single canonical issue cleanly represents the whole set. If forced, issue 44534 is the broadest and most central report in the v5 non-persistent-buffer regression family, but the cluster is too mixed for it to be a reliable duplicate anchor.", + "summary": "This cluster is highly heterogeneous: the issues cover unrelated models, tokenizers, config loading, dtype/device mismatches, and warning/message regressions. None of the soft-similarity pairs look like the same underlying bug, so all are rejected as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue is safe here; the items are not one duplicate family but many unrelated bugs across different subsystems and models.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44534 is the best global representative only in a loose sense because it is broad and clearly framed; however, the overall set is too diverse for any issue to be a strong canonical for the entire cluster.", + "best_issue_reason": "No global best issue for deduping stands out, because the soft links are mostly false positives and no item can represent the whole set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:39692", - "right": "issue:45834", - "accept": false, - "reason": "Different products and failure modes: SigLIP2 doc example/model-processor mismatch and quantization errors vs Kosmos2.5 OCR long-input index error." - }, { "left": "issue:43643", "right": "issue:44877", "accept": false, - "reason": "Both involve config loading, but one is missing fields from trust_remote_code AutoConfig and the other is strict config blocking granite_speech; not the same bug." + "reason": "Both are config-loading failures, but one is missing fields with trust_remote_code and the other is strict granite_speech config loading; different bugs." }, { "left": "issue:43335", "right": "issue:43756", "accept": false, - "reason": "Unrelated model-specific architecture issues: sparse-layer creation in SwitchTransformers vs RoPE layer dropping in Smollm3." + "reason": "SwitchTransformers sparse-layer creation and Smollm3 RoPE-layer dropping are unrelated model-architecture bugs." }, { - "left": "issue:45987", - "right": "issue:46932", + "left": "issue:44448", + "right": "issue:44671", "accept": false, - "reason": "StaticCache seq length contract bug is unrelated to Gemma4 quantized vision/audio embedder dtype casting." + "reason": "Pegasus v4/v5 output drift and CamemBERT masked-LM prediction regressions affect different models and code paths." }, { - "left": "issue:44448", - "right": "issue:44671", + "left": "issue:45987", + "right": "issue:46932", "accept": false, - "reason": "Both are Transformers v5 regressions, but one is Pegasus output drift and the other is CamemBERT masked LM prediction regression; different code paths." + "reason": "StaticCache seq-length return type and Gemma4 quantized embedder dtype casting are different cache vs quantization issues." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "Tokenizer crash in LayoutLMv2 NER/padding is unrelated to AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + "reason": "LayoutLMv2 tokenization crashes and AutoTokenizer loading FastSpeech2ConformerTokenizer fail in different tokenizer paths." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Mistral-3 image-preprocessor dtype mismatch and Gemma3n variable-name collision are unrelated failures." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Segfault loading Llava Next models is not the same as an MPS histogram kernel missing for Int." + "reason": "Llava Next loading segfaults and MPS histogram_mps int errors are different subsystems and error modes." }, { "left": "issue:44991", "right": "issue:46291", "accept": false, - "reason": "Tokenizer load failure for a specific model and pipeline import failure due kernels version requirements are different issues." - }, - { - "left": "issue:43618", - "right": "issue:43697", - "accept": false, - "reason": "CLIPOutput attentions assignment and RTDetrV2 output drift are separate v5 regressions with different affected components." - }, - { - "left": "issue:43716", - "right": "issue:44008", - "accept": false, - "reason": "Different multimodal dtype/forward bugs: image preprocessor dtype mismatch vs Gemma3n variable-name collision in forward()." + "reason": "Tokenizer loading for est-roberta and kernels version/revision import errors are unrelated." }, { "left": "issue:46294", "right": "issue:46463", "accept": false, - "reason": "Video chat-template token/feature mismatch in Qwen3.5-9B is unrelated to T5 decoder input validation messaging." + "reason": "Qwen3.5-9B video chat-template mismatch and T5 decoder-input validation are different model-specific bugs." }, { "left": "issue:44805", "right": "issue:44936", "accept": false, - "reason": "Mask shape IndexError and trainer.evaluate failing after train are not the same underlying bug." + "reason": "Mask/index shape errors and trainer.evaluate() failing after train are not the same underlying issue." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Missing CLIPOutput attentions and RTDetrV2 output drift are distinct regressions." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Device mismatch during CPU offload for bark-small is not the same as bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel." + "reason": "CPU offload device mismatch in bark-small and bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are different problems." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty-transcription failure after align_special_tokens and Kosmos2.5 long OCR index error are unrelated." + "reason": "Whisper empty-transcription failure after align_special_tokens and Kosmos2.5 long-OCR index errors do not share a concrete code path." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "torch.compile SDPA failure in Bamba-9B-v2 is not the same as dtype mismatch inference failures." + "reason": "torch.compile/SDPA breakage in Bamba-9B-v2 is unrelated to dtype mismatch inference bugs." }, { - "left": "issue:43421", - "right": "issue:44811", + "left": "issue:44479", + "right": "issue:45834", "accept": false, - "reason": "Runtime post-processor update bug in TokenizersBackend is distinct from Whisper batch_decode ignoring skip_special_tokens." + "reason": "Qwen VL video-input regression and Kosmos2.5 OCR index errors are different model/input pipelines." }, { - "left": "issue:43577", - "right": "issue:43818", + "left": "issue:43782", + "right": "issue:44977", "accept": false, - "reason": "BLIP2 dtype propagation and Video-LLaVA missing temporal attention/shared weights are unrelated model bugs." + "reason": "Both mention Qwen, but one is a weight_only loading error and the other is a flash-attention generation bug; not the same defect." }, { - "left": "issue:43644", - "right": "issue:43749", + "left": "issue:45183", + "right": "issue:45479", "accept": false, - "reason": "Non-persistent buffer junk in v5 and FSDP CPU RAM efficient loading failure are different regressions with different symptoms and fixes." + "reason": "Input validation messaging across models and single-label-classification zero-loss behavior are different bug classes." }, { - "left": "issue:45563", - "right": "issue:47276", + "left": "issue:46752", + "right": "issue:47337", "accept": false, - "reason": "Both are warning-message issues, but one is about num_return_sequences in paged generate and the other about synced_gpus in continuous batching; not the same change." + "reason": "apply_chat_template empty-list crash and bbox-coordinate flaky IndexError are unrelated." }, { - "left": "issue:43582", - "right": "issue:46512", + "left": "issue:43644", + "right": "issue:43749", "accept": false, - "reason": "Apple Silicon TypeError in caching_allocator_warmup is unrelated to stale config.dtype after from_config(dtype=...)." + "reason": "Non-persistent buffer junk filling and FSDP CPU RAM efficient loading are separate v5 regressions with different code paths." }, { - "left": "issue:44610", - "right": "issue:45072", + "left": "issue:45563", + "right": "issue:47276", "accept": false, - "reason": "Processor output-size mismatch for OmDet-Turbo is unrelated to dtype mismatch inference bugs." + "reason": "A stale num_return_sequences warning and a misleading synced_gpus warning are different warning-message issues." }, { - "left": "issue:45183", - "right": "issue:45479", + "left": "issue:43582", + "right": "issue:46512", "accept": false, - "reason": "Input validation XOR error messaging and single_label_classification zero-loss behavior are different sequence-classification problems." + "reason": "Apple Silicon caching_allocator_warmup TypeError and from_config dtype/config staleness are unrelated." }, { - "left": "issue:46752", - "right": "issue:47337", + "left": "issue:43577", + "right": "issue:43818", "accept": false, - "reason": "apply_chat_template empty-list crash and bbox coordinate range flaky test do not share the same code path." + "reason": "BLIP2 dtype propagation and Video-LLaVA temporal-attention bugs affect different models and implementations." }, { - "left": "issue:43782", - "right": "issue:44977", + "left": "issue:43421", + "right": "issue:44811", "accept": false, - "reason": "Qwen3VL from_pretrained weight_only error and Qwen3.5 flash-attention generation failure are separate model/runtime issues." + "reason": "Special-token post-processor updates and Whisper batch_decode skip_special_tokens handling are different tokenizer/processor behaviors." }, { - "left": "issue:44534", - "right": "issue:47458", + "left": "issue:44610", + "right": "issue:45072", "accept": false, - "reason": "Non-persistent buffers being filled with junk is not the same as GPT2 c_proj weights being incorrectly initialised." + "reason": "Processor image-size mismatch and bfloat16 dtype mismatch are not the same bug." }, { - "left": "issue:43644", - "right": "issue:44534", + "left": "issue:44534", + "right": "issue:47458", "accept": false, - "reason": "These are closely related v5 regression reports, but they are still not clearly the same concrete bug from the available evidence; keep them separate unless a later fix directly links them." + "reason": "Non-persistent buffer junk and GPT2 c_proj weight initialization are both initialization-related but not the same concrete issue." } ] }, @@ -6801,49 +6885,49 @@ "issue:47337" ], "soft_pairs": [ - "issue:43540|issue:45381", "issue:34634|issue:36331", "issue:28218|issue:29159", + "issue:43540|issue:45381", "issue:43720|issue:45072", "issue:43645|issue:46665", - "issue:47276|issue:47337", "issue:43749|issue:44291", + "issue:47276|issue:47337", + "issue:43742|issue:45020", + "issue:42175|issue:45081", "issue:43931|issue:46846", "issue:43825|issue:43994", - "issue:42175|issue:45081", - "issue:43742|issue:45020", "issue:43388|issue:46032", "issue:47254|issue:47333", - "issue:46697|issue:46728", "issue:42915|issue:47333", + "issue:46697|issue:46728", "issue:44671|issue:45701", - "issue:47254|issue:47332", "issue:43782|issue:45325", "issue:43577|issue:43582", + "issue:47254|issue:47332", "issue:44589|issue:45042", - "issue:46851|issue:47337", "issue:36010|issue:43645", "issue:43262|issue:47217", + "issue:46851|issue:47337", "issue:43262|issue:43819", - "issue:43653|issue:44912", "issue:43582|issue:44857", - "issue:45003|issue:46285", - "issue:36010|issue:36331" + "issue:43653|issue:44912", + "issue:36010|issue:36331", + "issue:45003|issue:46285" ], - "prepared_review_unit_hash": "28c22fa232e25cabd83f72eb5ecf861bb76bdafb8dc4bdc3c9f2e05903114f0a", + "prepared_review_unit_hash": "f76093db5cd51352d606fd8d5c239a4483e319520842f93c60a23dad11adead7", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36275, + "serialized_chars": 36276, "estimated_input_tokens": 9069, "estimated_eval_tokens": 18394 }, @@ -6852,180 +6936,180 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous. None of the soft pairs look like the same concrete bug/change closely enough to merge, so I rejected all candidate duplicate edges. The only borderline cases are broad thematic overlaps (tokenizers, loading, VL models), but they still describe different failures and code paths.", - "confidence": 0.89, - "canonical_issue_reason": "No true duplicate group stands out in this cluster. If one issue must serve as a representative, #45020 is the broadest and most general regression report, but it is not a duplicate of the others.", + "summary": "The cluster is highly heterogeneous and does not look like a duplicate set overall. Most soft pairs are only superficially similar (same subsystem or version-regression framing) but point to different code paths, models, or error modes.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45701 is the broadest tokenizer-regression report in the set and the best fit for the small tokenizer-related subset, but the cluster is not a coherent duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "#45020 is the most broadly scoped and useful representative issue in this set, with a clear regression framing and wide impact; however, it still does not subsume the other reports.", + "best_issue_reason": "45701 is the most general issue here; it best represents the only clearly overlapping theme (tokenization/version behavior), though it is still not a true canonical for the full cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43540", - "right": "issue:45381", - "accept": false, - "reason": "Both are VL/video-related, but they describe different model families and different failure modes; not the same concrete bug." - }, { "left": "issue:34634", "right": "issue:36331", "accept": false, - "reason": "Unrelated subsystems: BarkProcessor voice preset vs Trainer compute_loss signature." + "reason": "Different features and failures: BarkProcessor voice preset handling vs CustomTrainer.compute_loss API mismatch." }, { "left": "issue:28218", "right": "issue:29159", "accept": false, - "reason": "Tokenizer behavior overlaps, but one is about an added-token spacing bug and the other is a broader slow/fast tokenizer inconsistency; too vague to treat as the same defect." + "reason": "Both concern tokenizers, but the described bugs differ: extra space after added tokens vs slow/fast tokenizer inconsistency." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both are video-related, but they affect different Qwen models and different failure modes (ValueError vs wrong position_ids)." }, { "left": "issue:43720", "right": "issue:45072", "accept": false, - "reason": "Different loading/inference issues affecting different models and paths; not the same bug." + "reason": "Unrelated load/inference bugs: BitNet packed-weight unpacking vs dtype mismatches in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:43645", "right": "issue:46665", "accept": false, - "reason": "Both involve model construction, but one is a Jupyter/custom-model initialization breakage and the other is from_pretrained behavior inside Python classes." - }, - { - "left": "issue:47276", - "right": "issue:47337", - "accept": false, - "reason": "A warning-message issue and a flaky bbox-range test are unrelated." + "reason": "Both mention from_pretrained/model initialization, but one is a Jupyter custom-model regression and the other is behavior inside Python classes; not the same bug." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Both touch loading, but FSDP CPU RAM efficient loading and init_empty_weights argument handling are distinct code-path failures." + "reason": "Different loading paths and root causes: FSDP CPU RAM efficient loading vs init_empty_weights getting an unexpected _is_hf_initialized argument." }, { - "left": "issue:43931", - "right": "issue:46846", + "left": "issue:47276", + "right": "issue:47337", "accept": false, - "reason": "Different models and different root causes; one is a weight-shape mismatch, the other a tensor-parallel plan omission." + "reason": "Completely unrelated: a warning message issue vs a flaky bounding-box test failure." }, { - "left": "issue:43825", - "right": "issue:43994", + "left": "issue:43742", + "right": "issue:45020", "accept": false, - "reason": "Pipeline error-message wording and SigLIP2 bad outputs are different issues." + "reason": "A model-specific key error is not the same as the broad remote_code breakage; different underlying failures." }, { "left": "issue:42175", "right": "issue:45081", "accept": false, - "reason": "Backend packaging/install issue vs tokenizer regex patch crash; unrelated." + "reason": "TensorFlow backend packaging vs Mistral regex patch crash; no shared code-path bug." }, { - "left": "issue:43742", - "right": "issue:45020", + "left": "issue:43931", + "right": "issue:46846", "accept": false, - "reason": "Both are loading regressions, but the concrete failures differ: MobileLLM key error vs remote_code breakage." + "reason": "Different Qwen3-VL loading/shape issues and different mechanics (weight mismatch vs tensor-parallel plan omissions)." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message wording vs incorrect SigLIP2 outputs; one is messaging, the other is model behavior." }, { "left": "issue:43388", "right": "issue:46032", "accept": false, - "reason": "Metric gathering label truncation is unrelated to Mamba2Mixer cache correctness." + "reason": "Different training/metrics and generation-cache bugs; they do not share the same underlying code path." }, { "left": "issue:47254", "right": "issue:47333", "accept": false, - "reason": "Gradient-checkpointing/ZeRO checkpoint errors are not the same as disk-offload save failures." - }, - { - "left": "issue:46697", - "right": "issue:46728", - "accept": false, - "reason": "Both are image-processing shape issues, but over-padding and patch-count calculation are different bugs in different processors." + "reason": "Both involve training/offloading, but one is checkpointing with PEFT+DeepSpeed and the other is disk-offload save failure." }, { "left": "issue:42915", "right": "issue:47333", "accept": false, - "reason": "FP8 quantization failure and offloaded model save failure are different problems." + "reason": "Different Qwen models and different problems: FP8 config failure vs offloaded model save failure." }, { - "left": "issue:44671", - "right": "issue:45701", + "left": "issue:46697", + "right": "issue:46728", "accept": false, - "reason": "A masked-LM prediction regression and a broad tokenization version-change report are not the same concrete defect." + "reason": "Both are image-processing bugs, but one is over-padding and the other is patch-count calculation; not the same defect." }, { - "left": "issue:47254", - "right": "issue:47332", + "left": "issue:44671", + "right": "issue:45701", "accept": false, - "reason": "Different offloading-related failures: checkpointing error vs cannot save disk-offloaded model." + "reason": "A CamemBERT masked-LM regression is not the same as a generic tokenizer-version change report." }, { "left": "issue:43782", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen VL issues, but one is a weight_only loading error and the other is a rope/position-id bug." + "reason": "Different Qwen2.5-VL issues: weight_only loading error vs rope index scaling for still images." }, { "left": "issue:43577", "right": "issue:43582", "accept": false, - "reason": "Completely different areas: dtype propagation in BLIP2 vs Apple Silicon allocator warmup TypeError." + "reason": "Blip2 dtype propagation and Apple Silicon allocator warmup are unrelated." }, { - "left": "issue:44589", - "right": "issue:45042", + "left": "issue:47254", + "right": "issue:47332", "accept": false, - "reason": "Float8 storage lookup failure and PIL backend torchvision requirement are unrelated." + "reason": "Different offloading/training bugs: gradient checkpointing errors vs inability to save disk-offloaded Qwen3-VL." }, { - "left": "issue:46851", - "right": "issue:47337", + "left": "issue:44589", + "right": "issue:45042", "accept": false, - "reason": "Variable-length position_ids in Qwen3.5 is unrelated to a flaky bbox-coordinate test." + "reason": "Different domains: missing float8 storage type vs PIL image processors unnecessarily requiring torchvision." }, { "left": "issue:36010", "right": "issue:43645", "accept": false, - "reason": "ImportError for GenerationMixin and Jupyter custom-model initialization failures are different symptoms and code paths." + "reason": "Both are import/loading regressions, but one is GenerationMixin import failure and the other is notebook-defined custom model initialization." }, { "left": "issue:43262", "right": "issue:47217", "accept": false, - "reason": "Audio chat-template sampling-rate defaulting and VL image-placeholder failures are unrelated." + "reason": "Audio chat-template sampling-rate defaulting and VL image-placeholder failures are different multimodal bugs." }, { - "left": "issue:43262", - "right": "issue:43819", + "left": "issue:46851", + "right": "issue:47337", "accept": false, - "reason": "Audio processor sampling-rate default and DAC latent/forward mismatch are unrelated." + "reason": "Variable-length position_id handling is unrelated to the bounding-box flaky test." }, { - "left": "issue:43653", - "right": "issue:44912", + "left": "issue:43262", + "right": "issue:43819", "accept": false, - "reason": "Tokenizer special-token registration bug and MXFP4 quantization fallback are unrelated." + "reason": "Different model/pipeline bugs: audio processor chat template defaults vs DAC.from_latents mismatch." }, { "left": "issue:43582", "right": "issue:44857", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError and LwDetrImageLoss AMP crash are unrelated." + "reason": "Apple Silicon allocator TypeError is unrelated to LwDetrImageLoss AMP crash." }, { - "left": "issue:45003", - "right": "issue:46285", + "left": "issue:43653", + "right": "issue:44912", "accept": false, - "reason": "Unsafe sys.modules access and CLIP text-model structure inconsistency are different implementation issues." + "reason": "BigBirdTokenizer special-token registration and MXFP4 quantization fallback are unrelated." }, { "left": "issue:36010", "right": "issue:36331", "accept": false, - "reason": "Import path regression and Trainer compute_loss signature mismatch are unrelated." + "reason": "ImportError for GenerationMixin and CustomTrainer compute_loss signature mismatch are separate API breakages." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is not the same as the CLIPTextModel structural inconsistency." } ] }, @@ -7099,231 +7183,234 @@ "issue:47050" ], "soft_pairs": [ + "issue:43531|issue:45042", "issue:46175|issue:46932", "issue:45372|issue:46326", - "issue:43531|issue:45042", "issue:43550|issue:44912", "issue:42175|issue:46291", "issue:43638|issue:44936", - "issue:43825|issue:43827", + "issue:42915|issue:44821", "issue:43819|issue:43927", "issue:43257|issue:43329", "issue:43452|issue:44991", "issue:45081|issue:45406", - "issue:42915|issue:44821", - "issue:44060|issue:44368", - "issue:43526|issue:43697", "issue:43257|issue:43696", + "issue:43825|issue:43827", + "issue:43526|issue:43697", + "issue:44060|issue:44368", "issue:43854|issue:44977", "issue:45859|issue:47050", - "issue:43906|issue:46209", - "issue:43746|issue:44811", + "issue:45084|issue:45538", "issue:28370|issue:29159", + "issue:43906|issue:46209", + "issue:42915|issue:43742", "issue:38175|issue:44610", - "issue:45084|issue:45538", + "issue:43746|issue:44811", "issue:43329|issue:44479", - "issue:42915|issue:43742", - "issue:42490|issue:43257", + "issue:44291|issue:45042", "issue:43262|issue:44811", - "issue:44291|issue:45042" + "issue:42490|issue:43257" ], - "prepared_review_unit_hash": "3deede376aa08c4962e9ef4e63e32c6aecdc66bdf33a153bfbb71a5999527a89", + "prepared_review_unit_hash": "ce340248d0e80cbfb7e912699fab3f42bc167ef0c0e2ebf9a61ced9257f6726a", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36205, - "estimated_input_tokens": 9052, - "estimated_eval_tokens": 18360 + "serialized_chars": 36203, + "estimated_input_tokens": 9051, + "estimated_eval_tokens": 18358 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No true duplicate cluster emerged: the items cover many unrelated Transformers bugs/regressions across tokenizers, multimodal processors, quantization, loading, and docs. All soft edges look like topical similarities rather than the same underlying change or bug. No PRs are present.", - "confidence": 0.92, - "canonical_issue_reason": "No clear canonical issue; the set spans unrelated regressions and none of the soft pairs is a true duplicate.", + "summary": "The cluster is mostly heterogeneous one-off issues. Only one soft duplicate looks plausible: the Qwen video-input regression pair. Everything else appears to be separate bugs or unrelated subsystems.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44479 is the best umbrella report for the only clear duplicate pair: it describes the broader v5.3.0 video-input regression across multiple Qwen VL/MoE models, while issue:43329 reads like a narrower root-cause report in the multimodal token-count path.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue best represents the whole set for deduplication; #43825 is the closest to a general anchor, but it only covers pipeline messaging/docs and not the rest.", + "best_issue_reason": "issue:44479 is the strongest standalone issue in the cluster because it is broad, regression-framed, and captures the user-visible impact across several models.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Different bugs: Qwen3-MoE sliding-window logic vs PIL image processors incorrectly requiring torchvision." + }, { "left": "issue:46175", "right": "issue:46932", "accept": false, - "reason": "Both are Gemma4-related, but one is a huge vision patch-embedder memory issue and the other is dtype casting for quantized embedders; different code paths and fixes." + "reason": "Same model family, but different code paths: a huge one-hot position-embedding allocation vs quantized vision/audio dtype casting." }, { "left": "issue:45372", "right": "issue:46326", "accept": false, - "reason": "Different failures: Gemma4 processor loading breaks on a missing import, while the other is a mistral_common continue_final_message bug." - }, - { - "left": "issue:43531", - "right": "issue:45042", - "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and PIL/torchvision dependency requirements are unrelated bugs." + "reason": "Both involve mistral_common, but one is an import error during Gemma 4 processor loading and the other is continue_final_message tokenizer behavior." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "One is a torch.compile/SDPA failure in Bamba, the other is MXFP4 loading falling back to bf16 for git-oss-20b." + "reason": "Unrelated issues: torch.compile/SDPA failure for Bamba vs MXFP4 quantization fallback for git-oss-20b." }, { "left": "issue:42175", "right": "issue:46291", "accept": false, - "reason": "Packaging missing TensorFlow in a torch extra is not the same as kernels v0.15 import-time version/revision validation." + "reason": "Different dependency failures: missing TensorFlow in a torch install vs kernels v0.15 import-time version/revision requirements." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "Different training issues: DeepSpeed ZeRO-3 with a non-pretrained Bert model vs evaluate() failing after train()." + "reason": "Both are training-related, but one is a ZeRO-3/non-pretrained Bert IndexError and the other is evaluate() failing after train(); not the same bug." }, { - "left": "issue:43825", - "right": "issue:43827", + "left": "issue:42915", + "right": "issue:44821", "accept": false, - "reason": "Related to pipeline v5 removals, but one is a runtime error message and the other is documentation drift; not the same bug." + "reason": "Completely different subsystems: Qwen3-MoE FP8 config failure vs AutoImageProcessor loading from URL." }, { "left": "issue:43819", "right": "issue:43927", "accept": false, - "reason": "DAC from_latents/STE mismatch and DiaConfig token ID persistence are separate model/config bugs." + "reason": "Different models and failures: DAC latent reconstruction mismatch vs DiaConfig losing custom token IDs after save/load." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Accelerate+DeepSpeed Qwen3-MoE weight conversion and undefined multimodal token helper names are distinct issues." + "reason": "One is Qwen3-MoE weight conversion with accelerate+deepspeed; the other is a multimodal token-count video-branch bug." }, { "left": "issue:43452", "right": "issue:44991", "accept": false, - "reason": "Both concern loading, but gguf_file support and a tokenizer regression for EMBEDDIA/est-roberta are different root causes." + "reason": "Both concern tokenizer/model loading, but one is gguf_file support for auto loaders and the other is a specific v5 tokenizer regression for EMBEDDIA/est-roberta." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Mistral regex patching crash and Gemma4Processor missing _tokenizer are unrelated loading failures." + "reason": "Both are loading-time errors, but they affect different components: Mistral tokenizer regex patching vs Gemma4Processor tokenization state." }, { - "left": "issue:42915", - "right": "issue:44821", + "left": "issue:43257", + "right": "issue:43696", "accept": false, - "reason": "FineGrainedFP8Config failure for Qwen3Moe and AutoImageProcessor URL loading are unrelated." + "reason": "No shared underlying bug: Qwen3-MoE conversion issue vs GPT-oss-20b CUDA OOM." }, { - "left": "issue:44060", - "right": "issue:44368", + "left": "issue:43825", + "right": "issue:43827", "accept": false, - "reason": "Both mention tied-weights warnings, but they involve different models and different mis-tied parameters." + "reason": "Related to pipeline removal, but one is a runtime error message and the other is stale docs; not the same concrete bug/change." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output drift in v5 are separate vision bugs." + "reason": "Different image/model bugs: BeitImageProcessorFast label reduction vs RTDetrV2 output divergence in v5." }, { - "left": "issue:43257", - "right": "issue:43696", + "left": "issue:44060", + "right": "issue:44368", "accept": false, - "reason": "Qwen3-MoE weight conversion and GPT-oss-20b CUDA OOM are not the same problem." + "reason": "Both mention tied-weights warnings, but the concrete symptoms and models differ; not clearly the same code path." }, { "left": "issue:43854", "right": "issue:44977", "accept": false, - "reason": "GLM-4.7-Flash unit-test loading and Qwen3.5 flash-attention generation issues are different model-specific failures." + "reason": "Different model-specific inference/loading problems: GLM-4.7-Flash unit-test loading vs Qwen3.5 flash-attention generation failure." }, { "left": "issue:45859", "right": "issue:47050", "accept": false, - "reason": "CPU offload incompatibility in Qwen3_5MoeTextRotaryEmbedding is unrelated to Canine ONNX export dtype mismatch." + "reason": "Unrelated bugs: CPU-offload incompatibility in Qwen3_5Moe rotary embedding vs CanineModel fp16 ONNX attention dtype mismatch." }, { - "left": "issue:43906", - "right": "issue:46209", + "left": "issue:45084", + "right": "issue:45538", "accept": false, - "reason": "One is just an isolated reproduction of another issue; the other is an FP8 inference failure on B200, not the same bug." + "reason": "No overlap beyond being bugs: template compilation TypeError vs CLIPTokenizer model_max_length sentinel value." }, { - "left": "issue:43746", - "right": "issue:44811", + "left": "issue:28370", + "right": "issue:29159", "accept": false, - "reason": "GraniteSpeech PEFT local checkpoint loading and Whisper batch_decode special-token handling are different code paths." + "reason": "Both are tokenizer-related, but one is missing vocab_file for custom SentencePiece models and the other is slow/fast tokenizer inconsistency in general." }, { - "left": "issue:28370", - "right": "issue:29159", + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "Different reports: an isolated reproduction of an earlier issue vs an FP8 inference failure on B200 hardware." + }, + { + "left": "issue:42915", + "right": "issue:43742", "accept": false, - "reason": "Missing vocab_file for custom SentencePiece models is a specific tokenizer-save/load bug, not the same as general slow/fast tokenizer inconsistency." + "reason": "Different model-loading failures with no shared code path: Qwen3-MoE FP8 config vs MobileLLM-125M key error." }, { "left": "issue:38175", "right": "issue:44610", "accept": false, - "reason": "Unexpected zero probabilities in SigLIP2 and OmDet-Turbo processor size mismatch are distinct vision-model regressions." + "reason": "Both are vision-model complaints, but one is zero probabilities in SigLIP2 and the other is an image-size mismatch in OmDet-Turbo." }, { - "left": "issue:45084", - "right": "issue:45538", + "left": "issue:43746", + "right": "issue:44811", "accept": false, - "reason": "Template compilation errors and CLIPTokenizer model_max_length are unrelated tokenizer/template issues." + "reason": "Different functionality: PEFT adapter loading for GraniteSpeech vs Whisper batch_decode skip_special_tokens handling." }, { "left": "issue:43329", "right": "issue:44479", - "accept": false, - "reason": "The latter is a broad video-input regression across Qwen VL variants; the former is a specific undefined-name bug in one multimodal-token helper, so they are not clearly the same fix." - }, - { - "left": "issue:42915", - "right": "issue:43742", - "accept": false, - "reason": "Qwen3Moe FP8 config failure and MobileLLM-125M loading key error are unrelated model-loading problems." + "accept": true, + "reason": "These look like the same video-input regression: both are about broken multimodal/video handling in Qwen VL/MoE models, and 43329 provides a likely root cause in the video token-count path while 44479 is the broader regression report." }, { - "left": "issue:42490", - "right": "issue:43257", + "left": "issue:44291", + "right": "issue:45042", "accept": false, - "reason": "Trailing-slash local path import warning and Qwen3-MoE accelerate/deepspeed conversion failure are different issues." + "reason": "Unrelated issues: init_empty_weights/_is_hf_initialized argument handling vs PIL image processors requiring torchvision." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio processor chat-template sampling-rate defaulting and Whisper batch_decode skip_special_tokens are separate audio-processing bugs." + "reason": "Different audio/processor bugs: apply_chat_template sampling-rate default vs Whisper batch_decode special-token skipping." }, { - "left": "issue:44291", - "right": "issue:45042", + "left": "issue:42490", + "right": "issue:43257", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized signature mismatch and PIL backend torchvision dependency requirements are unrelated." + "reason": "No shared bug: local-path trailing-slash import warning vs Qwen3-MoE weight conversion with accelerate+deepspeed." } ] }, - "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 pair, issue:43329 vs issue:44479, is plausibly the same Qwen video-input regression at different levels of specificity. The other pairwise rejections are appropriately cautious and do not overclaim duplication." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -7355,6 +7442,7 @@ "issue:43454", "issue:43525", "issue:43526", + "issue:43531", "issue:43540", "issue:43577", "issue:43643", @@ -7380,11 +7468,10 @@ "issue:45759", "issue:45803", "issue:46291", - "issue:46421", - "issue:46424", "issue:46701", "issue:47020", "issue:47145", + "issue:47332", "issue:47333", "issue:47458" ], @@ -7392,221 +7479,234 @@ "issue:43643|issue:45759", "issue:43329|issue:45600", "issue:39692|issue:44912", - "issue:43526|issue:44554", - "issue:43329|issue:43450", "issue:44387|issue:44617", "issue:34567|issue:40990", + "issue:43526|issue:44554", + "issue:43329|issue:43450", "issue:44589|issue:47333", "issue:29159|issue:34567", "issue:45081|issue:47020", + "issue:43425|issue:46291", "issue:43257|issue:43931", + "issue:44291|issue:44821", "issue:43756|issue:47458", - "issue:43425|issue:46291", "issue:44987|issue:45538", "issue:43525|issue:46291", "issue:44912|issue:47145", - "issue:44291|issue:44821", - "issue:45459|issue:45803", "issue:43540|issue:46701", "issue:43742|issue:44291", - "issue:43650|issue:43906", + "issue:45459|issue:45803", + "issue:43454|issue:43577", "issue:43749|issue:44756", + "issue:43531|issue:44821", "issue:28282|issue:34689", - "issue:28282|issue:29870", - "issue:43454|issue:43577", + "issue:43650|issue:43906", "issue:33357|issue:34567", - "issue:46421|issue:46424" + "issue:28282|issue:29870", + "issue:44589|issue:47332" ], - "prepared_review_unit_hash": "19ec37957571c8ba536cff4bf9049a494adcb15596118981961c12c2e7b5dc82", + "prepared_review_unit_hash": "4fd66c353fb7957d466f257224a4c112f2441b9a5a38f714cfa3568f74ac5b25", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35346, - "estimated_input_tokens": 8837, - "estimated_eval_tokens": 17930 + "soft_pair_count": 28, + "serialized_chars": 35606, + "estimated_input_tokens": 8902, + "estimated_eval_tokens": 18060 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous: most items are distinct bugs in different models, loaders, tokenizers, cache paths, or training utilities. None of the soft pairs look like the same underlying defect, so I would not collapse any of them as duplicates.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is a false-positive mix of unrelated Transformers issues spanning config loading, tokenizers, video/multimodal processing, quantization, training state, and model-specific regressions. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.89, + "canonical_issue_reason": "If forced to pick a single anchor, issue 43643 is the closest because it is a broad config-loading regression, but it does not actually subsume the rest of the cluster, so there is no real canonical duplicate here.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 43643 is the most representative standalone report in this set, but only as a loose loader/config example; it is not a good global duplicate center for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "Both involve config/model-loading behavior, but one is about missing fields from `AutoConfig.from_pretrained(trust_remote_code=True)` and the other about `AutoModelForCausalLM.from_config` not unwrapping `text_config` in composite multimodal configs. Different code paths and failures." + "reason": "Related config-loading surface, but different bugs: missing returned fields from AutoConfig vs not unwrapping text_config in AutoModelForCausalLM." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "One is a video multimodal token-counting bug with undefined helpers; the other is stale `auto_mappings.py` references breaking CI. Same broad area only, not the same issue." + "reason": "One is a video token counting undefined-variable bug; the other is stale auto_mappings references to removed configs." }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 docs/model-processor mismatch plus quantization failure is unrelated to GPT-OSS MXFP4 quantization falling back to bf16. Different model families and root causes." + "reason": "Documentation example errors are unrelated to an MXFP4 quantization fallback bug." }, { - "left": "issue:43526", - "right": "issue:44554", + "left": "issue:44387", + "right": "issue:44617", "accept": false, - "reason": "`reduce_labels` returning the wrong label count in `BeitImageProcessorFast` is unrelated to an MPS attention correctness bug. No shared code path." + "reason": "Both are memory-related, but one is an int4 reserved-memory regression and the other is a model-specific CUDA OOM." }, { - "left": "issue:43329", - "right": "issue:43450", + "left": "issue:34567", + "right": "issue:40990", "accept": false, - "reason": "Both mention video processing, but one is undefined multimodal-token helper names and the other is incorrect batched video input shape. Different failure modes." + "reason": "Trainer token-count bookkeeping is unrelated to high perplexity on a benchmark." }, { - "left": "issue:44387", - "right": "issue:44617", + "left": "issue:43526", + "right": "issue:44554", "accept": false, - "reason": "Both mention OOM, but one is increased CUDA reserved memory under int4 quantization and the other is a Sam3Video memory blowup. Not the same bug." + "reason": "Image-processor label reduction bug and MPS attention correctness issue are different code paths and failures." }, { - "left": "issue:34567", - "right": "issue:40990", + "left": "issue:43329", + "right": "issue:43450", "accept": false, - "reason": "Trainer token-count tracking not updating is unrelated to high perplexity on GPT-OSS. Different subsystem and symptom." + "reason": "Both mention video processing, but one is an undefined helper in multimodal token counting and the other is a batched-shape bug in processors." }, { "left": "issue:44589", "right": "issue:47333", "accept": false, - "reason": "Float8 storage lookup failure during loading is unrelated to saving a disk-offloaded Qwen3 model. Different storage/load path." + "reason": "Storage lookup failure and disk-offload save failure are distinct loading/saving problems." }, { "left": "issue:29159", "right": "issue:34567", "accept": false, - "reason": "Tokenizer slow/fast inconsistency is not the same as `num_input_tokens_seen` failing to update in `TrainerState`." + "reason": "Tokenizer slow/fast inconsistency is unrelated to TrainerState token accounting." }, { "left": "issue:45081", "right": "issue:47020", "accept": false, - "reason": "Both are tokenizer issues, but one is a Mistral regex patch crashing on `backend_tokenizer` and the other is XLMRoberta passing a dict where `Unigram` expects a sequence. Different models and bugs." + "reason": "Different tokenizer internals: backend_tokenizer attribute access vs passing the wrong type to Unigram." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch version incompatibility and kernels package import requirements are separate dependency issues." }, { "left": "issue:43257", "right": "issue:43931", "accept": false, - "reason": "Both involve Qwen3 loading, but one is MOE weights not converted with accelerate+deepspeed and the other is shape mismatch for Qwen3-VL-30B-A3B-Instruct. Not the same defect." + "reason": "MoE weight conversion under accelerate/deepspeed is not the same as a Qwen3-VL weight-shape mismatch." }, { - "left": "issue:43756", - "right": "issue:47458", + "left": "issue:44291", + "right": "issue:44821", "accept": false, - "reason": "Smollm3 RoPE-layer handling and GPT-2 `c_proj` initialization are unrelated model-initialization bugs." + "reason": "init_empty_weights argument regression is unrelated to loading AutoImageProcessor from a URL." }, { - "left": "issue:43425", - "right": "issue:46291", + "left": "issue:43756", + "right": "issue:47458", "accept": false, - "reason": "Torch 2.10 incompatibility is a dependency/version issue; kernels v0.15 import failure is a separate package-contract problem." + "reason": "RoPE layer-count/config mismatch is not the same as GPT2 c_proj initialization being wrong." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Loading `physical-intelligence/fast` on transformers>=5.1.0 and CLIPTokenizer's huge `model_max_length` are unrelated issues." + "reason": "Version-specific model loading failure and CLIPTokenizer max_length misuse are different bugs." }, { "left": "issue:43525", "right": "issue:46291", "accept": false, - "reason": "Missing `pad_token_id` on `Llama4Config` is a model config bug, not the `kernels` version/revision import problem." + "reason": "Missing pad_token_id on Llama4Config is unrelated to pipelines import failing on kernels metadata." }, { "left": "issue:44912", "right": "issue:47145", "accept": false, - "reason": "Both are quantization/loading related, but one falls back to bf16 for MXFP4 and the other misses `weight_scale` for frozen FP8 checkpoints. Different code paths." - }, - { - "left": "issue:44291", - "right": "issue:44821", - "accept": false, - "reason": "`init_empty_weights` receiving an unexpected `_is_hf_initialized` arg is unrelated to loading `AutoImageProcessor` from a URL." - }, - { - "left": "issue:45459", - "right": "issue:45803", - "accept": false, - "reason": "Both are error-handling concerns, but one hides tokenizer decode errors without protobuf and the other uses a bare `except` in `FuyuBatchFeature.convert_to_tensors()`. Different locations and symptoms." + "reason": "MXFP4 fallback-to-bf16 loading bug and FP8 frozen-checkpoint weight_scale omission are different quantization issues." }, { "left": "issue:43540", "right": "issue:46701", "accept": false, - "reason": "Both are multimodal pipeline failures, but one is Qwen3OmniMoe video input validation and the other is DiffusionGemma image-text-to-text output handling. Not the same issue." + "reason": "Qwen3OmniMoe video-input ValueError is unrelated to DiffusionGemmaGenerationOutput pipeline behavior." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "KeyError loading MobileLLM-125M is unrelated to the `_is_hf_initialized` argument regression in `init_empty_weights`." + "reason": "MobileLLM key error is model-specific; init_empty_weights is a broader API regression." }, { - "left": "issue:43650", - "right": "issue:43906", + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Both involve exception handling, but they are different functions and different failure modes." + }, + { + "left": "issue:43454", + "right": "issue:43577", "accept": false, - "reason": "`ADD THE DATA` is not a technical bug report; it is unrelated to the isolated reproduction of issue 38071." + "reason": "Weight tying causing garbage generation is unrelated to BLIP2 dtype propagation staying float32." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "FSDP CPU RAM-efficient loading being broken and disabling mmap on Strix Halo to avoid OOM are separate memory/loading problems." + "reason": "FSDP CPU RAM-efficient loading and disabling mmap on Strix Halo are separate memory-management issues." }, { - "left": "issue:28282", - "right": "issue:34689", + "left": "issue:43531", + "right": "issue:44821", "accept": false, - "reason": "PyTorch-not-found import error is a missing dependency issue; Llama 3.2 Vision model loading breakage in 4.46.2 is a specific model compatibility regression." + "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor URL loading are unrelated." }, { "left": "issue:28282", - "right": "issue:29870", + "right": "issue:34689", "accept": false, - "reason": "AutoModel import failing without PyTorch and Bart generate hitting an unbound decoder attention mask are unrelated." + "reason": "Missing PyTorch installation error is not the same as a model-loading regression for Llama 3.2 Vision." }, { - "left": "issue:43454", - "right": "issue:43577", + "left": "issue:43650", + "right": "issue:43906", "accept": false, - "reason": "AyaVision tying `lm_head` weights and BLIP2 dtype staying float32 are both model-loading quirks, but they affect different models and mechanisms." + "reason": "A vague placeholder issue is not a duplicate of a concrete reproduction report." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "MacOS bus error with CLIP community model is unrelated to TrainerState token counting." + "reason": "MacOS bus error in CLIP loading is unrelated to TrainerState token counting." }, { - "left": "issue:46421", - "right": "issue:46424", + "left": "issue:28282", + "right": "issue:29870", + "accept": false, + "reason": "ImportError for missing torch and BART generate decoder_attention_mask bug are different failures." + }, + { + "left": "issue:44589", + "right": "issue:47332", + "accept": false, + "reason": "Float8 storage lookup failure and disk-offloaded model save failure are separate code paths." + }, + { + "left": "issue:43749", + "right": "issue:47333", "accept": false, - "reason": "Both touch cache logic, but one is lazy cache init causing recompilation and the other is `max_cache_length` only being honored for Quantized Cache. Different behaviors." + "reason": "FSDP CPU RAM efficient loading and saving a disk-offloaded Qwen3 model are different offloading problems." } ] }, @@ -7646,7 +7746,6 @@ "issue:42915", "issue:43404", "issue:43452", - "issue:43531", "issue:43540", "issue:43582", "issue:43697", @@ -7654,11 +7753,10 @@ "issue:43824", "issue:44062", "issue:44479", + "issue:44488", "issue:44554", "issue:44560", "issue:44568", - "issue:44589", - "issue:44821", "issue:44857", "issue:44898", "issue:44936", @@ -7672,6 +7770,8 @@ "issue:45405", "issue:45488", "issue:45538", + "issue:46421", + "issue:46424", "issue:46693", "issue:46821", "issue:46832", @@ -7681,250 +7781,250 @@ ], "soft_pairs": [ "issue:44479|issue:46693", - "issue:44589|issue:47332", "issue:42175|issue:43824", "issue:44568|issue:45356", - "issue:43531|issue:44821", "issue:42915|issue:45538", - "issue:24643|issue:29159", + "issue:46421|issue:46424", "issue:43582|issue:47254", - "issue:43697|issue:45357", + "issue:24643|issue:29159", + "issue:44568|issue:45488", "issue:44936|issue:46821", - "issue:45137|issue:45405", + "issue:43697|issue:45357", "issue:43404|issue:45127", + "issue:45538|issue:46869", "issue:28282|issue:29127", + "issue:45137|issue:45405", + "issue:41720|issue:43540", "issue:44062|issue:44991", "issue:44857|issue:46832", - "issue:44568|issue:45488", - "issue:41720|issue:43540", - "issue:45538|issue:46869", - "issue:39379|issue:44554", "issue:43540|issue:44479", "issue:43540|issue:44560", + "issue:39379|issue:44554", "issue:42915|issue:43749", "issue:42915|issue:47332", + "issue:33666|issue:35308", "issue:44479|issue:45381", - "issue:34567|issue:36331", "issue:43452|issue:43582", + "issue:34567|issue:36331", "issue:27957|issue:28282", - "issue:33666|issue:35308", "issue:44062|issue:45081", - "issue:38175|issue:44898" + "issue:38175|issue:44898", + "issue:42175|issue:44488" ], - "prepared_review_unit_hash": "9cbb5f668013eb2f9a051a24f088c9da5972de6cf61c66453ba463df12c71585", + "prepared_review_unit_hash": "b96053d58aab1dbc773c08a24bcb9a4d4d28e7d5c4ed10f436fe25559116a195", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36708, - "estimated_input_tokens": 9177, - "estimated_eval_tokens": 18610 + "serialized_chars": 36761, + "estimated_input_tokens": 9191, + "estimated_eval_tokens": 18638 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous and does not form one duplicate set. The only convincing soft edge is the Qwen-VL video-input regression pair (44479 \u2194 45381); the rest span unrelated tokenizer, training, packaging, and model-specific bugs.", - "confidence": 0.69, - "canonical_issue_reason": "issue:44479 is the broadest and most central anchor in the only real overlap area: it describes a Qwen-VL video-input regression affecting multiple model variants, which fits the narrower qwen2.5-vl video-position bug as a manifestation of the same problem.", + "summary": "Mostly unrelated issue cluster. I only found one credible duplicate edge: the Qwen2.5-VL video-input regression report aligns with the broader Qwen video-input regression issue. The rest look like separate bugs across different subsystems/models.", + "confidence": 0.89, + "canonical_issue_reason": "issue:44479 is the best anchor because it states the broad video-input regression affecting multiple Qwen multimodal variants, and it cleanly subsumes the more specific qwen2.5-vl video-position report.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best issue to represent the cluster\u2019s closest duplicate-like theme because it has the widest scope and is the clearest umbrella for the Qwen video-input regression subgroup.", + "best_issue_reason": "issue:44479 is the strongest representative issue in the cluster: it has the clearest regression framing, the broadest model scope, and matches the one accepted duplicate edge.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44479", "right": "issue:46693", "accept": false, - "reason": "Both mention regressions, but one is Qwen-VL video input and the other is FA2 long-context generation performance; different code paths and symptoms." - }, - { - "left": "issue:44589", - "right": "issue:47332", - "accept": false, - "reason": "Different failure modes: missing Float8 storage vs disk-offloaded Qwen3-VL save issues; not the same bug." + "reason": "Both mention video/long-context performance, but one is a Qwen multimodal video-input regression and the other is an FA2 long-context slowdown. Different bugs and code paths." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "Packaging/backend dependency issue vs missing model import; only broad install/import context overlaps." + "reason": "Backend packaging vs missing-model import error; unrelated failure modes." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Both are tokenizer regressions, but they target different models and different tokenizer mechanics (special tokens vs codec/fix_mistral_regex handling)." - }, - { - "left": "issue:43531", - "right": "issue:44821", - "accept": false, - "reason": "Qwen3-MoE sliding-window/model bug and AutoImageProcessor URL loading are unrelated." + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different code paths. Too broad to be the same bug." }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "FineGrainedFP8Config failure in Qwen3MoE is unrelated to CLIPTokenizer max-length behavior." + "reason": "Qwen3Moe FP8 failure and CLIP tokenizer max-length handling are unrelated." }, { - "left": "issue:24643", - "right": "issue:29159", + "left": "issue:46421", + "right": "issue:46424", "accept": false, - "reason": "DeepSpeed training weight shape error and slow/fast tokenizer inconsistency are unrelated subsystems." + "reason": "Both are cache-related, but one is lazy init causing recompilation and the other is max_cache_length being ignored outside quantized cache. Separate issues." }, { "left": "issue:43582", "right": "issue:47254", "accept": false, - "reason": "Apple Silicon allocator TypeError and PEFT/ZeRO-3 checkpointing error are distinct training/runtime failures." + "reason": "Different distributed-training failures: Apple Silicon warmup TypeError vs PEFT/DeepSpeed checkpointing error." }, { - "left": "issue:43697", - "right": "issue:45357", + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed training runtime error and tokenizer consistency are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45488", "accept": false, - "reason": "RTDetr output mismatch and Qwen3.5 visual encoder save_pretrained key corruption are different bugs." + "reason": "Both involve tokenizers, but one is special-token duplication and the other is LlamaTokenizer pre-tokenizer overriding. Different underlying bugs." }, { "left": "issue:44936", "right": "issue:46821", "accept": false, - "reason": "Trainer evaluate-after-train failure and TimesFM window_size AttributeError do not share a concrete code path." + "reason": "Trainer lifecycle bug vs TimesFM window_size AttributeError; not the same problem." }, { - "left": "issue:45137", - "right": "issue:45405", + "left": "issue:43697", + "right": "issue:45357", "accept": false, - "reason": "DeepSpeed ZeRO-3 deque error and PEFT version bump are unrelated." + "reason": "RT-DETR output mismatch and Qwen3.5 save_pretrained key corruption are different model behaviors and code paths." }, { "left": "issue:43404", "right": "issue:45127", "accept": false, - "reason": "Mistral3 weight tying bug and LoRA merge collapse with extended vocab are different model-state issues." - }, - { - "left": "issue:28282", - "right": "issue:29127", - "accept": false, - "reason": "Environment-missing-PyTorch import error and LayoutLMv3 box validation clarity issue are not duplicates." + "reason": "Both mention tied weights, but one is a specific lm_head tying bug and the other is LoRA merge collapse with extended vocab. Not the same concrete defect." }, { - "left": "issue:44062", - "right": "issue:45081", + "left": "issue:45538", + "right": "issue:46869", "accept": false, - "reason": "AddedToken kwarg collision and Mistral regex patch crash are different tokenizer bugs." + "reason": "Tokenizer max_length issue vs YOLO multi-GPU device-map failure; unrelated." }, { - "left": "issue:44857", - "right": "issue:46832", + "left": "issue:28282", + "right": "issue:29127", "accept": false, - "reason": "LwDetr AMP crash and RT-DETR backbone/features mismatch are separate detection-model failures." + "reason": "Missing PyTorch import error is unrelated to LayoutLMv3 box-validation messaging." }, { - "left": "issue:44568", - "right": "issue:45488", + "left": "issue:45137", + "right": "issue:45405", "accept": false, - "reason": "Both concern tokenizer behavior, but the affected models and failure mechanisms differ substantially." + "reason": "DeepSpeed ZeRO-3 deque error and PEFT version bump are unrelated." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Qwen3 auto device mapping cuda assert and Qwen3OmniMoe video processing are unrelated." + "reason": "Both are Qwen model failures, but one is auto device mapping CUDA assert and the other is video-input processing in Qwen3OmniMoe. Different bugs." }, { - "left": "issue:45538", - "right": "issue:46869", + "left": "issue:44062", + "right": "issue:44991", "accept": false, - "reason": "Tokenizer max-length constant and YOLO multi-GPU device-map failure are unrelated." + "reason": "Generic AddedToken constructor conflict vs a specific tokenizer-loading failure for EMBEDDIA/est-roberta." }, { - "left": "issue:39379", - "right": "issue:44554", + "left": "issue:44857", + "right": "issue:46832", "accept": false, - "reason": "mllama mask handling and MPS attention correctness are different attention-path issues." + "reason": "Different object-detection bugs: AMP/CUDA loss crash vs RT-DETR forward shape mismatch." }, { "left": "issue:43540", "right": "issue:44479", "accept": false, - "reason": "Both involve video input, but they are different model families and the reported failures point to different code paths." + "reason": "Both concern Qwen video inputs, but 43540 is a specific Qwen3OmniMoe ValueError while 44479 is a broader multi-model regression. Not enough evidence they are the same concrete bug." }, { "left": "issue:43540", "right": "issue:44560", "accept": false, - "reason": "Qwen3OmniMoe video error and Qwen3-vl-embedding StopIteration are not the same concrete bug." + "reason": "Both are video-input failures, but they involve different Qwen model families and different errors." + }, + { + "left": "issue:39379", + "right": "issue:44554", + "accept": false, + "reason": "Mllama mask handling and MPS attention correctness are unrelated." }, { "left": "issue:42915", "right": "issue:43749", "accept": false, - "reason": "Qwen3MoE FP8 config failure and FSDP CPU RAM efficient loading breakage are unrelated training/config issues." + "reason": "Qwen3Moe FP8 config failure and FSDP CPU RAM efficient loading are different distributed/runtime issues." }, { "left": "issue:42915", "right": "issue:47332", "accept": false, - "reason": "FineGrainedFP8Config failure and disk-offloaded save failure are different Qwen3-family issues." + "reason": "FP8 config failure and disk-offload save failure are unrelated." + }, + { + "left": "issue:33666", + "right": "issue:35308", + "accept": false, + "reason": "Both are multi-GPU training issues, but one is Qwen2-VL-specific and the other is iterable-dataset length mismatch. Different root causes." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "Both describe the same Qwen2.5-VL video-input regression in 5.3.0; 45381 is a narrower symptom of the broader video-position bug." - }, - { - "left": "issue:34567", - "right": "issue:36331", - "accept": false, - "reason": "TrainerState token counting and custom compute_loss signature mismatch are distinct trainer API issues." + "reason": "Strong match: both describe the same Qwen2.5-VL video-input regression in transformers 5.3.0, with 45381 giving the specific vision_position_ids symptom." }, { "left": "issue:43452", "right": "issue:43582", "accept": false, - "reason": "gguf_file loading breakage and Apple Silicon allocator TypeError are unrelated." + "reason": "gguf_file tokenizer/model loading breakage is unrelated to the Apple Silicon allocator warmup TypeError." }, { - "left": "issue:27957", - "right": "issue:28282", + "left": "issue:34567", + "right": "issue:36331", "accept": false, - "reason": "XLMRoberta Flash Attention 2 and missing PyTorch import are completely different problems." + "reason": "TrainerState token counting and custom compute_loss signature mismatch are different trainer API issues." }, { - "left": "issue:33666", - "right": "issue:35308", + "left": "issue:27957", + "right": "issue:28282", "accept": false, - "reason": "Qwen2-VL multi-GPU training and IterableDataset length mismatch are both training-related but not the same bug." + "reason": "Flash Attention support for XLM-Roberta is unrelated to missing PyTorch import detection." }, { "left": "issue:44062", "right": "issue:45081", "accept": false, - "reason": "Duplicate of the earlier tokenizer-pair verdict: the errors are unrelated." + "reason": "AddedToken keyword collision and Mistral regex patch crashing on backend_tokenizer are different tokenizer bugs." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Siglip2 zero probabilities and Perceiver non-default resolution failure are separate model-specific issues." + "reason": "Siglip2 zero probabilities and Perceiver interpolation failure are unrelated vision-model issues." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "TensorFlow packaging/backend omission and a specific tokenizer load failure are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The cluster summary is grounded in the packet and stays conservative: it correctly treats the set as mostly heterogeneous while elevating the Qwen video-input regression pair as the only plausible duplicate-like edge. The accepted 44479 \u2194 45381 link is supported by the titles, and the other verdicts read as appropriately cautious rather than overstated." + "feedback": "The summary is well grounded in the packet and stays conservative. The lone accepted edge (44479 \u2194 45381) is clearly supported by the issue titles, and the rejected soft edges are framed as distinct bugs without overclaiming. No mergeability or PR-pair concerns here." }, "evaluator_used": true, "retried": false, @@ -7949,8 +8049,6 @@ "issue:28282", "issue:28370", "issue:29870", - "issue:42175", - "issue:42796", "issue:43257", "issue:43452", "issue:43493", @@ -7968,6 +8066,7 @@ "issue:44625", "issue:44737", "issue:44792", + "issue:44821", "issue:44849", "issue:44877", "issue:44964", @@ -7988,6 +8087,7 @@ "issue:46512", "issue:46693", "issue:46701", + "issue:46821", "issue:46899", "issue:47276", "issue:47332", @@ -7995,252 +8095,248 @@ "issue:47405" ], "soft_pairs": [ - "issue:42175|issue:44488", + "issue:43742|issue:44589", "issue:43493|issue:43994", "issue:43577|issue:46512", - "issue:43742|issue:44589", "issue:28282|issue:28370", - "issue:44792|issue:44964", + "issue:45672|issue:45674", "issue:43742|issue:47333", - "issue:44987|issue:45907", - "issue:43756|issue:45325", + "issue:44792|issue:44964", "issue:43716|issue:46701", "issue:44849|issue:45701", - "issue:43756|issue:46399", - "issue:45672|issue:45674", + "issue:44987|issue:45907", + "issue:43756|issue:45325", "issue:45823|issue:46463", - "issue:43257|issue:43756", - "issue:46463|issue:47276", + "issue:43756|issue:46399", "issue:44737|issue:46463", + "issue:46463|issue:47276", + "issue:43257|issue:43756", "issue:45405|issue:45600", "issue:43906|issue:46693", - "issue:43540|issue:46693", - "issue:43257|issue:47405", "issue:44625|issue:44877", + "issue:43540|issue:46693", + "issue:44466|issue:45005", "issue:43742|issue:47332", - "issue:44568|issue:46491", "issue:43577|issue:45072", + "issue:43257|issue:47405", + "issue:44568|issue:46491", "issue:43452|issue:44488", - "issue:44466|issue:45005", "issue:24643|issue:29870", "issue:46463|issue:46899", - "issue:42796|issue:43906" + "issue:44821|issue:46821" ], - "prepared_review_unit_hash": "104d04022dfb4911441aed8b92f133e381b0140cb2e0e93d3cfba04c49781764", + "prepared_review_unit_hash": "451a11cea3f8d5838425b716d2952ae0d5c884f2b11e55056297ee597f6b4659", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36960, - "estimated_input_tokens": 9240, - "estimated_eval_tokens": 18736 + "soft_pair_count": 29, + "serialized_chars": 36624, + "estimated_input_tokens": 9156, + "estimated_eval_tokens": 18568 }, "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 bugs: tokenization, dtype propagation, config loading, multimodal preprocessing, offloading, and performance regressions. The soft pairs share broad subsystems or similar words, but not the same concrete failure path, so I reject all edges. No single canonical issue/PR stands out cleanly.", - "confidence": 0.83, - "canonical_issue_reason": "No single canonical issue: the items span unrelated regressions (tokenization, dtype, configs, multimodal, offloading, performance).", + "summary": "This cluster is mostly heterogeneous model-loading / dtype / tokenizer bug reports, with only two plausible duplicate mini-pairs: SigLIP2 output-quality issues and tied-weights handling in v5. Most other soft matches are superficial subsystem overlap rather than the same bug.", + "confidence": 0.66, + "canonical_issue_reason": "issue:44466 is the clearest, most root-cause-oriented report in the only broader duplicate family here: v5 tied-weight serialization varying by device. It is the best anchor for the tied-weights subgroup.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44466 is the clearest self-contained regression report and the strongest representative of the small tied-weights serialization theme.", + "best_issue_reason": "issue:44466 is the strongest overall representative of the cluster\u2019s limited duplicate signal: it names the concrete tied-weights bug directly and is broad enough to subsume the closely related v5 tied-weights report.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42175", - "right": "issue:44488", + "left": "issue:43742", + "right": "issue:44589", "accept": false, - "reason": "One is a TensorFlow packaging/backend omission; the other is a model-loading failure. Different bug classes." + "reason": "Different failures: model loading key error vs Float8 storage lookup/serialization error. Same broad area, not the same bug." }, { "left": "issue:43493", "right": "issue:43994", - "accept": false, - "reason": "Same model family, but the reported symptoms and likely fixes differ; no clear shared code-path bug from the titles alone." + "accept": true, + "reason": "Both point to SigLIP2 producing wrong outputs versus expected behavior/reference implementation; likely the same underlying model implementation bug." }, { "left": "issue:43577", "right": "issue:46512", "accept": false, - "reason": "Both involve dtype state, but one is BLIP2-specific model/qformer loading and the other is generic from_config dtype staleness. Not the same bug." - }, - { - "left": "issue:43742", - "right": "issue:44589", - "accept": false, - "reason": "A key error while loading a model and a missing Float8 storage type error are unrelated failures." + "reason": "Related dtype theme, but one is BLIP2/Qformer staying float32 while the other is from_config leaving config.dtype stale. Different code paths." }, { "left": "issue:28282", "right": "issue:28370", "accept": false, - "reason": "Import-time PyTorch dependency error vs tokenizer attribute handling; different layers and different fixes." + "reason": "ImportError for missing PyTorch is unrelated to missing vocab_file for custom SentencePiece models." }, { - "left": "issue:44792", - "right": "issue:44964", + "left": "issue:45672", + "right": "issue:45674", "accept": false, - "reason": "Different models and different failure modes: a Janus test regression versus a Phi-4 multimodal load regression." + "reason": "Both involve 4-bit quantization, but one is an unsafe dtype cast in Gemma4 audio modules and the other is a skip-module/default-exclusion assertion bug." }, { "left": "issue:43742", "right": "issue:47333", "accept": false, - "reason": "They both mention loading/saving, but the models and failure paths differ; not enough to treat as the same underlying bug." - }, - { - "left": "issue:44987", - "right": "issue:45907", - "accept": false, - "reason": "One is a model-load regression, the other is an API forwarding bug in list_repo_files. Unrelated." + "reason": "Different model families and different operations: loading MobileLLM vs saving a disk-offloaded Qwen3 model." }, { - "left": "issue:43756", - "right": "issue:45325", + "left": "issue:44792", + "right": "issue:44964", "accept": false, - "reason": "Both touch positional encoding/rope concepts, but one drops layers and the other mis-scales temporal positions. Different concrete bugs." + "reason": "Both are multimodal-model issues, but they target different models and different failure surfaces (test generation vs loading)." }, { "left": "issue:43716", "right": "issue:46701", "accept": false, - "reason": "Image preprocessor dtype mismatch vs a pipeline output/processing failure; different code paths." + "reason": "Different models and bugs: Mistral-3 preprocessor dtype mismatch vs DiffusionGemma pipeline failure." }, { "left": "issue:44849", "right": "issue:45701", "accept": false, - "reason": "Hidden-states behavior in Qwen3.5 vs tokenizer-version-dependent tokenization changes. Unrelated." + "reason": "Qwen3.5 hidden-state bug is unrelated to tokenization changes across versions." }, { - "left": "issue:43756", - "right": "issue:46399", + "left": "issue:44987", + "right": "issue:45907", "accept": false, - "reason": "RoPE layer truncation and missing kernel registration decorator are not the same bug." + "reason": "Model loading regression vs repo-files revision forwarding API bug; not the same underlying change." }, { - "left": "issue:45672", - "right": "issue:45674", + "left": "issue:43756", + "right": "issue:45325", "accept": false, - "reason": "Both are 4-bit inference issues, but one is an unsafe uint8 finfo cast and the other is lm_head exclusion handling. Different root causes." + "reason": "Both mention positional/RoPE logic, but they are different model-specific bugs with different symptoms and fixes." }, { "left": "issue:45823", "right": "issue:46463", "accept": false, - "reason": "Device-map/PLE mismatch in Gemma4 vs a T5 decoder-input validation message bug. Different issues." + "reason": "Gemma4 device placement mismatch and T5 decoder input validation messaging are unrelated." }, { - "left": "issue:43257", - "right": "issue:43756", + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "Different model subsystems: Smollm3 RoPE-layer count bug vs Qwen3.5 kernel registration failure." + }, + { + "left": "issue:44737", + "right": "issue:46463", "accept": false, - "reason": "Qwen3 MOE conversion under accelerate+deepspeed vs Smollm3 rope-layer dropping. Different models and problems." + "reason": "XLNet CPU allocation bug is unrelated to T5 decoder validation error handling." }, { "left": "issue:46463", "right": "issue:47276", "accept": false, - "reason": "An incorrect error message and a misleading warning message are separate message-level issues, not the same bug." + "reason": "Both are message/warning quality issues, but they are different code paths and not the same bug." }, { - "left": "issue:44737", - "right": "issue:46463", + "left": "issue:43257", + "right": "issue:43756", "accept": false, - "reason": "XLNet CPU tensor placement bug vs T5 validation messaging. No shared code path." + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed is unrelated to Smollm3 RoPE layer dropping." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "A dependency version bump/release issue vs an auto_mappings CI break from removed configs. Different maintenance problems." + "reason": "One is a version-pin/dependency issue; the other is stale auto_mappings entries breaking CI." }, { "left": "issue:43906", "right": "issue:46693", "accept": false, - "reason": "One is a reproduction of an older issue; the other is a FA2 long-context performance regression. Not the same change or bug." + "reason": "Insufficient evidence of the same bug; one is a reproduction of another issue, the other is a performance regression in FA2 long-context generation." }, { - "left": "issue:43540", - "right": "issue:46693", + "left": "issue:44625", + "right": "issue:44877", "accept": false, - "reason": "Video-input ValueError in Qwen3OmniMoe vs a generation performance regression. Different failures." + "reason": "Both are config-related, but one is Qwen3.5 num_labels propagation and the other is strict config loading for granite_speech." }, { - "left": "issue:43257", - "right": "issue:47405", + "left": "issue:43540", + "right": "issue:46693", "accept": false, - "reason": "Conversion failure during loading with deepspeed vs silent wrong-class conversion skipping. Related theme, but distinct bugs." + "reason": "Video-input handling in Qwen3OmniMoe is unrelated to the FA2 performance regression." }, { - "left": "issue:44625", - "right": "issue:44877", - "accept": false, - "reason": "Both involve config handling, but one is num_labels propagation and the other is strict config loading. Not the same root cause." + "left": "issue:44466", + "right": "issue:45005", + "accept": true, + "reason": "Both describe v5 tied-weights handling problems; the shared underlying issue is serialization/weight-tie behavior changing in v5 for translation-style models." }, { "left": "issue:43742", "right": "issue:47332", "accept": false, - "reason": "Model key error on load vs disk-offloaded save failure. Different bugs." - }, - { - "left": "issue:44568", - "right": "issue:46491", - "accept": false, - "reason": "Special-token insertion regression vs leading-space encode/decode regression. Both tokenizer-related, but not the same issue." + "reason": "Different model-loading vs disk-offload save failure; no same code-path evidence." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Both are dtype mismatches, but they affect different models and code paths; too broad to merge." + "reason": "Both are dtype bugs, but they affect different model stacks and likely different fixes." }, { - "left": "issue:43452", - "right": "issue:44488", + "left": "issue:43257", + "right": "issue:47405", "accept": false, - "reason": "Both concern loading models, but the specific mechanisms differ and there is no clear shared failing path from the titles alone." + "reason": "Deepspeed conversion failure is not the same as silent conversion skipping when loading through the wrong Auto class." }, { - "left": "issue:44466", - "right": "issue:45005", + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Tokenizer regressions, but different behaviors and different models; not the same bug." + }, + { + "left": "issue:43452", + "right": "issue:44488", "accept": false, - "reason": "Both mention tied weights, but one is a device-dependent serialization regression and the other is a translation-model tied-weight behavior report. Too little evidence to merge." + "reason": "Both are loading-related, but one is a gguf_file regression across AutoTokenizer/AutoModelForCausalLM and the other is a specific model load failure without shared code-path evidence." }, { "left": "issue:24643", "right": "issue:29870", "accept": false, - "reason": "DeepSpeed weight-shape error vs Bart generate decoder-mask initialization bug. Unrelated." + "reason": "DeepSpeed training weight-shape error is unrelated to Bart generate() using decoder_attention_mask before assignment." }, { "left": "issue:46463", "right": "issue:46899", "accept": false, - "reason": "T5 decoder validation messaging vs Gemma4 multimodal 4-bit dtype casting. Different code paths and fixes." + "reason": "T5 validation messaging and Gemma4 4-bit multimodal dtype casting are unrelated bugs." }, { - "left": "issue:42796", - "right": "issue:43906", + "left": "issue:44821", + "right": "issue:46821", "accept": false, - "reason": "Tokenizer BPE-marker handling vs a reproduction issue reference. No shared bug." + "reason": "AutoImageProcessor URL-loading failure is unrelated to TimesFM window_size AttributeError." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays appropriately conservative. The two accepted soft-edge pairs are plausible from the issue titles alone, and the many rejections are well-justified by differing models, failure surfaces, or code paths. No overstatement stands out." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -8257,11 +8353,13 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ + "issue:33666", + "issue:36010", "issue:38175", "issue:40444", + "issue:42796", "issue:43295", "issue:43452", - "issue:43525", "issue:43572", "issue:43643", "issue:43646", @@ -8271,6 +8369,7 @@ "issue:43720", "issue:43749", "issue:43761", + "issue:43906", "issue:43927", "issue:44361", "issue:44442", @@ -8278,7 +8377,6 @@ "issue:44737", "issue:44756", "issue:44811", - "issue:44821", "issue:44843", "issue:44857", "issue:44898", @@ -8297,8 +8395,6 @@ "issue:46121", "issue:46170", "issue:46178", - "issue:46701", - "issue:46821", "issue:46829", "issue:46846", "issue:47000", @@ -8307,244 +8403,238 @@ "issue:47404" ], "soft_pairs": [ + "issue:42796|issue:43906", "issue:43452|issue:44843", - "issue:44821|issue:46821", - "issue:43697|issue:46829", - "issue:43720|issue:44898", - "issue:44756|issue:47404", "issue:45663|issue:46846", - "issue:44913|issue:46121", + "issue:44756|issue:47404", + "issue:43720|issue:44898", + "issue:43697|issue:46829", "issue:44737|issue:45588", + "issue:44913|issue:46121", "issue:43653|issue:44361", "issue:44493|issue:46050", "issue:38175|issue:43295", "issue:45357|issue:45920", + "issue:44756|issue:47332", "issue:44361|issue:44442", "issue:46846|issue:47404", + "issue:40444|issue:44811", "issue:43697|issue:45216", - "issue:44756|issue:47332", - "issue:43688|issue:43749", "issue:44843|issue:44991", - "issue:40444|issue:44811", + "issue:43688|issue:43749", "issue:45372|issue:46170", - "issue:43572|issue:43643", "issue:44857|issue:47000", "issue:43761|issue:47013", - "issue:45542|issue:46178", + "issue:43572|issue:43643", "issue:43646|issue:43749", + "issue:45542|issue:46178", + "issue:33666|issue:36010", "issue:43688|issue:45701", "issue:43688|issue:45704", - "issue:43525|issue:46701", "issue:43653|issue:43927" ], - "prepared_review_unit_hash": "4dc0ab76db6a978404290dd40c0a0bae0da5a1afddd792f6500b3f28245b6980", + "prepared_review_unit_hash": "60691c649c04a7ea3122cf17d9e18271f63a9095be73541061bf08a0c45e4ff5", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36772, - "estimated_input_tokens": 9193, - "estimated_eval_tokens": 18642 + "serialized_chars": 36780, + "estimated_input_tokens": 9195, + "estimated_eval_tokens": 18646 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All 30 soft pairs look like false positives: they share broad themes like tokenizer loading, config regressions, or OOMs, but the concrete bugs, models, and code paths differ.", - "confidence": 0.82, - "canonical_issue_reason": "issue:45920 is the best umbrella issue here because it is a concrete, multi-model v5 tokenizer regression with clear user impact and a broader scope than the more model-specific reports.", + "summary": "This cluster is highly heterogeneous: the soft pairs mostly link unrelated bug reports that share only broad themes like tokenizers, loading, or regressions. I would not merge any of the candidate pairs.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45920 is the strongest anchor: it\u2019s a broad, high-impact v5 tokenizer regression affecting multiple models, with a clear symptom and wide applicability.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the strongest standalone issue: it is specific enough to reproduce, affects multiple model families, and reflects a real cross-model regression rather than a one-off symptom.", + "best_issue_reason": "issue:45920 is the best representative issue in this set because it is well-scoped, broadly relevant, and clearly describes a user-facing regression with multiple affected models.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:42796", + "right": "issue:43906", + "accept": false, + "reason": "Different tokenizer bugs/models; one is BPE marker handling, the other is an isolated reproduction of a different prior issue." + }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Both involve AutoTokenizer loading, but one is about gguf_file breaking model/tokenizer loading while the other is an offline model_info() call in _patch_mistral_regex." + "reason": "Both involve loading, but one is gguf_file handling and the other is an offline model_info call in _patch_mistral_regex." }, { - "left": "issue:44821", - "right": "issue:46821", + "left": "issue:45663", + "right": "issue:46846", "accept": false, - "reason": "Different failures: URL loading of AutoImageProcessor versus a TimesFM 2.5 window_size AttributeError." + "reason": "Different models and failure modes: FSDP2 KeyError in Gemma-4 vs Qwen3.5 tensor-parallel OOM/shape issues." }, { - "left": "issue:43697", - "right": "issue:46829", + "left": "issue:44756", + "right": "issue:47404", "accept": false, - "reason": "Different models and code paths: RTDetrV2 output divergence versus VideoPrism hidden_states being misassigned in ImageClassifierOutput." + "reason": "Both are OOM-related, but mmap on Strix Halo and destination-GPU contiguous() transients are different code paths." }, { "left": "issue:43720", "right": "issue:44898", "accept": false, - "reason": "BitNet packed-weight loading and Perceiver positional interpolation are unrelated bugs, despite both being model inference issues." + "reason": "Unrelated models and bugs; BitNet packed-weight loading vs Perceiver interpolation failure." }, { - "left": "issue:44756", - "right": "issue:47404", + "left": "issue:43697", + "right": "issue:46829", "accept": false, - "reason": "Both mention OOM, but the causes differ: mmap on Strix Halo versus unbudgeted contiguous() transients during device_map conversion." + "reason": "Both report output mismatches, but the models and output fields differ; not the same underlying defect." }, { - "left": "issue:45663", - "right": "issue:46846", + "left": "issue:44737", + "right": "issue:45588", "accept": false, - "reason": "Different architectures and bugs: Gemma-4 FSDP2 shared_kv_states reconstruction versus Qwen3.5 tensor-parallel plan omissions." + "reason": "XLNet CPU placement bug is unrelated to the flash_attention.py s_aux=None crash." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "One is rotary_pct not persisting on reload; the other is a TypeError converting rope params from a JSON-loaded list." - }, - { - "left": "issue:44737", - "right": "issue:45588", - "accept": false, - "reason": "XLNet CPU arange bug and flash_attention s_aux=None crash are unrelated implementation defects." + "reason": "Both touch rope/config logic, but one is rotary_pct persistence and the other is list-type validation in a helper." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "Both are tokenizer-related, but one is BigBird special-token registration and the other is an MLukeTokenizer AttributeError on tasks." + "reason": "Different tokenizer implementations and error modes: special-token decode issue vs MLuke AttributeError on tasks." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Unexpected position-id keys and sinq save/load breakage are different regression patterns and not the same underlying bug." + "reason": "Position-id warnings and quantized save/load breakage are not the same bug or code path." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Siglip2 zero probabilities and processor.tokenizer/image-to-tokenizer regression are separate multimodal issues." + "reason": "siglip2 zero probabilities and processor.tokenizer regression are separate vision/text issues." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Both concern Qwen3.5 / tokenization-related regressions, but one is incorrect visual encoder keys in save_pretrained while the other is wrong token IDs across several models." + "reason": "Qwen3.5 visual-encoder save keys and broad AutoTokenizer wrong-ID regression are different problems." + }, + { + "left": "issue:44756", + "right": "issue:47332", + "accept": false, + "reason": "Both are offloading-related, but mmap OOM avoidance and saving a disk-offloaded model are distinct failures." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Two different tokenizer failures: MLukeTokenizer task AttributeError versus AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + "reason": "Different tokenizer models and failures: MLuke task AttributeError vs FastSpeech2Conformer loading error." }, { "left": "issue:46846", "right": "issue:47404", "accept": false, - "reason": "Both can lead to OOM, but one is a tensor-parallel planning issue and the other is transient GPU memory from conversion ops during loading." - }, - { - "left": "issue:43697", - "right": "issue:45216", - "accept": false, - "reason": "Different model families and failure modes: RTDetrV2 output mismatch versus Qwen3.5 incorrect checkpoint save_pretrained behavior." + "reason": "Different subsystems: Qwen3.5 tensor-parallel plan vs device_map conversion-op OOM." }, { - "left": "issue:44756", - "right": "issue:47332", + "left": "issue:40444", + "right": "issue:44811", "accept": false, - "reason": "Both are memory/offloading related, but disable mmap for Strix Halo is not the same as failing to save a disk-offloaded Qwen3-VL model." + "reason": "IterableDataset multi-image finetuning failure is unrelated to Whisper batch_decode skip_special_tokens behavior." }, { - "left": "issue:43688", - "right": "issue:43749", + "left": "issue:43697", + "right": "issue:45216", "accept": false, - "reason": "Auxiliary-loss normalization in OLMoE/GPT Oss is unrelated to FSDP_CPU_RAM_EFFICIENT_LOADING being broken." + "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained checkpoint corruption are unrelated regressions." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer loading regressions, but the offline model_info() patch issue is distinct from EMBEDDIA/est-roberta failing on transformers >=5.0.0." + "reason": "Offline model_info access and est-roberta tokenizer loading regression do not share the same concrete code path." }, { - "left": "issue:40444", - "right": "issue:44811", + "left": "issue:43688", + "right": "issue:43749", "accept": false, - "reason": "Different subsystems: Qwen2.5-VL multi-image finetuning with IterableDataset versus Whisper batch_decode skip_special_tokens handling." + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are separate bugs." }, { "left": "issue:45372", "right": "issue:46170", "accept": false, - "reason": "One is a mistral_common import regression affecting Gemma 4 processors; the other is a Windows ROCm torch.distributed dependency problem." - }, - { - "left": "issue:43572", - "right": "issue:43643", - "accept": false, - "reason": "Both mention missing config fields, but one is StableLmConfig pad_token_idx after 5.0 and the other is a generic trust_remote_code AutoConfig field-loss bug." + "reason": "Gemma processor import failure and Windows ROCm torch.distributed dependency issue are different failures." }, { "left": "issue:44857", "right": "issue:47000", "accept": false, - "reason": "Different detection losses and errors: float16 AMP crash in LwDetrImageLoss versus infeasible cost matrix in RTDetrLoss." + "reason": "Different loss functions/models and different crash causes: AMP float16 failure vs infeasible cost matrix." }, { "left": "issue:43761", "right": "issue:47013", "accept": false, - "reason": "CLIPVisionModel hidden_states being None and xLSTM returning a bare tensor instead of a tuple are different forward-contract bugs." + "reason": "CLIP hidden_states regression and xLSTM return_last_states unpacking bug are unrelated." }, { - "left": "issue:45542", - "right": "issue:46178", + "left": "issue:43572", + "right": "issue:43643", "accept": false, - "reason": "Both are backend-registration issues, but tensorboard-without-TensorFlow and tensorflow_text missing from BACKENDS_MAPPING are distinct failures." + "reason": "StableLmConfig pad_token_idx loss and trust_remote_code missing fields are different config bugs." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Custom model initialization regression and FSDP CPU RAM efficient loading are unrelated issues." + "reason": "General custom model initialization regression vs FSDP CPU RAM loading; separate issues." }, { - "left": "issue:43688", - "right": "issue:45701", + "left": "issue:45542", + "right": "issue:46178", "accept": false, - "reason": "Auxiliary-loss normalization is unrelated to tokenization changing across transformers versions." + "reason": "Both are backend-related, but one is undefined tf backend from tensorboard-only install and the other is unsupported tensorflow_text backend." }, { - "left": "issue:43688", - "right": "issue:45704", + "left": "issue:33666", + "right": "issue:36010", "accept": false, - "reason": "Different concerns entirely: loss normalization versus T5 silently using apex.FusedRMSNorm and leaking memory." + "reason": "Qwen2-VL multi-GPU training and GenerationMixin import error are unrelated." }, { - "left": "issue:43525", - "right": "issue:46701", + "left": "issue:43688", + "right": "issue:45701", "accept": false, - "reason": "Llama4Config pad_token_id missing is a config attribute bug; DiffusionGemmaGenerationOutput pipeline failure is a different multimodal generation issue." + "reason": "Aux-loss normalization and tokenizer behavior changes across versions are separate bugs." }, { - "left": "issue:44843", - "right": "issue:44991", + "left": "issue:43688", + "right": "issue:45704", "accept": false, - "reason": "The offline model_info() regression and est-roberta tokenizer loading failure are not the same code-path problem." + "reason": "Aux-loss normalization has no concrete overlap with T5 apex.FusedRMSNorm memory leak." }, { "left": "issue:43653", "right": "issue:43927", "accept": false, - "reason": "BigBird special-token decode failure and DiaConfig losing custom token IDs after save/load are different tokenization/config persistence bugs." + "reason": "Both are tokenizer/config related, but BigBird special-token registration and DiaConfig custom token IDs are different issues." } ] }, @@ -8572,17 +8662,18 @@ "issue:32090", "issue:33666", "issue:34567", - "issue:36010", "issue:36331", + "issue:39692", "issue:41720", "issue:42898", "issue:43278", "issue:43295", + "issue:43334", "issue:43452", + "issue:43493", "issue:43504", "issue:43525", "issue:43550", - "issue:43572", "issue:43577", "issue:43606", "issue:43643", @@ -8596,21 +8687,20 @@ "issue:44464", "issue:44466", "issue:44479", - "issue:44554", "issue:44561", "issue:44568", "issue:44610", "issue:44625", - "issue:44734", "issue:44743", "issue:44977", "issue:45356", "issue:45405", "issue:45440", + "issue:45759", "issue:46050", "issue:46491", - "issue:46612", "issue:46693", + "issue:46701", "issue:46829", "issue:46832", "issue:46932", @@ -8618,259 +8708,246 @@ "issue:47276" ], "soft_pairs": [ - "issue:33666|issue:36010", + "issue:43525|issue:46701", "issue:43295|issue:45356", "issue:44479|issue:44625", "issue:44464|issue:44743", "issue:46832|issue:46932", "issue:43550|issue:44361", - "issue:43653|issue:44568", "issue:43761|issue:46829", + "issue:43653|issue:44568", "issue:41720|issue:44464", "issue:43452|issue:43577", "issue:33666|issue:36331", "issue:44977|issue:46693", "issue:28370|issue:34567", - "issue:44466|issue:46050", - "issue:43504|issue:43550", "issue:44464|issue:44977", + "issue:43504|issue:43550", + "issue:44466|issue:46050", + "issue:43782|issue:46693", "issue:42898|issue:46491", "issue:32090|issue:33666", - "issue:43782|issue:46693", "issue:45405|issue:47254", - "issue:43504|issue:43720", "issue:43606|issue:43927", "issue:43720|issue:43927", + "issue:43504|issue:43720", "issue:43643|issue:44561", - "issue:43278|issue:43688", "issue:43688|issue:45440", + "issue:43278|issue:43688", "issue:43504|issue:44610", + "issue:39692|issue:43493", "issue:44625|issue:47276", "issue:43720|issue:44361", - "issue:44734|issue:46612", - "issue:43577|issue:44554", - "issue:43525|issue:43572" + "issue:43334|issue:45759" ], - "prepared_review_unit_hash": "ec140a9cbe911ccd55f351838beaba2ac63060477bf60f68a607f8c1614ea4fb", + "prepared_review_unit_hash": "08164063efca208e52c9b9bd4ec6b20139609decf597ccdc4b8c3048acbedb21", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37571, - "estimated_input_tokens": 9393, - "estimated_eval_tokens": 19042 + "soft_pair_count": 31, + "serialized_chars": 37338, + "estimated_input_tokens": 9335, + "estimated_eval_tokens": 18926 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster does not form a true duplicate set: it mixes unrelated tokenizer, training, generation, config, and model-loading bugs. All soft links are only superficial topical similarity, so none should be merged.", - "confidence": 0.86, - "canonical_issue_reason": "No single canonical issue fits this cluster because the items are heterogeneous and do not describe the same underlying bug or change; they only share broad Transformers-related vocabulary.", + "summary": "These issues are a heterogeneous mix of unrelated Transformers bugs/regressions (tokenizers, model loading, training, generation, vision, MoE, quantization). None of the soft pairs look like the same underlying bug or change, so no duplicates are accepted.", + "confidence": 0.91, + "canonical_issue_reason": "Issue #46693 is the most central by discussion volume and inbound references, with a broad user-facing regression impact; it is the best single representative despite the overall cluster being heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "If one issue must be used as the anchor, issue:43295 is the strongest standalone report: it names a concrete regression, a specific version, and a reproducible failure mode in custom model code.", + "best_issue_reason": "#46693 is the strongest representative of the set because it is the most heavily discussed and clearly documented regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:33666", - "right": "issue:36010", + "left": "issue:43525", + "right": "issue:46701", "accept": false, - "reason": "Different problems: multi-GPU Qwen2-VL training vs a GenerationMixin import error. No shared code-path or fix." + "reason": "Different model families and failure modes: missing `pad_token_id` on Qwen3-VL config vs a DiffusionGemma image-text pipeline bug." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Both are regressions, but they affect different tokenizer/model paths and different failure modes." + "reason": "Both are regressions, but one is about processor/tokenizer access in custom model code and the other is about Kimi-K2.5 codec/tokenization handling; different code paths." }, { "left": "issue:44479", "right": "issue:44625", "accept": false, - "reason": "Same family (Qwen3.5) but unrelated bugs: video input regression vs config field propagation." + "reason": "Different bugs: video input regression across Qwen VL variants vs `num_labels` propagation into Qwen3.5 text config." }, { "left": "issue:44464", "right": "issue:44743", "accept": false, - "reason": "Both involve generation/caching, but the concrete bugs and model code paths are different." + "reason": "Both touch generation/caching, but one is chunked generation with compiled forward and the other is recurrent state reset in a modular Qwen3.5 file; not the same concrete bug." }, { "left": "issue:46832", "right": "issue:46932", "accept": false, - "reason": "RT-DETR feature-level handling and Gemma4 quantized dtype casting are unrelated issues." + "reason": "Unrelated areas: RT-DETR feature-level mismatch vs Gemma4 quantized embedder dtype casting." }, { "left": "issue:43550", "right": "issue:44361", "accept": false, - "reason": "These are unrelated: torch.compile/SDPA on Bamba vs a tokenizer AttributeError on tasks." + "reason": "Completely different subsystems: torch.compile + SDPA in Bamba vs tokenizer AttributeError in MLukeTokenizer." }, { - "left": "issue:43653", - "right": "issue:44568", + "left": "issue:43761", + "right": "issue:46829", "accept": false, - "reason": "Both concern special tokens, but the model-specific failures and root causes differ." + "reason": "Both involve hidden_states, but one is CLIPVisionModel omitting them and the other is VideoPrism classification output wiring; different implementation bugs." }, { - "left": "issue:43761", - "right": "issue:46829", + "left": "issue:43653", + "right": "issue:44568", "accept": false, - "reason": "Both touch output_hidden_states, but they are different model implementations and opposite output-shape issues." + "reason": "Both are tokenizer special-token issues, but they affect different tokenizers and different behaviors (mask token registration vs BOS/EOS insertion)." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert is unrelated to compiled-forward chunked generation inconsistency." + "reason": "Different failure modes: CUDA assert under auto device mapping for Qwen3 vs inconsistent outputs from compiled chunked generation." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "Tokenizer/model loading via gguf_file is unrelated to Blip2 dtype preservation." + "reason": "Different loading/serialization issues: gguf_file breaks AutoModel/AutoTokenizer loading vs Blip2 dtype not propagating to model/qformer." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Multi-GPU training failure and CustomTrainer.compute_loss signature mismatch are separate trainer bugs." + "reason": "Different trainer-related problems: Qwen2-VL multi-GPU training vs a custom `compute_loss` signature regression." }, { "left": "issue:44977", "right": "issue:46693", "accept": false, - "reason": "Both mention FA2, but one is a Qwen3.5 correctness bug and the other is a performance regression." + "reason": "Both mention generation and attention, but one is correctness for Qwen3.5 with flash-attention and the other is a performance regression with FA2 long-context generation." }, { "left": "issue:28370", "right": "issue:34567", "accept": false, - "reason": "SentencePiece vocab_file loading and TrainerState token counting are unrelated." + "reason": "Unrelated: SentencePiece `vocab_file` attribute handling vs TrainerState token accounting not updating." }, { - "left": "issue:44466", - "right": "issue:46050", + "left": "issue:44464", + "right": "issue:44977", "accept": false, - "reason": "Both involve save/load serialization, but the tied-weight/device issue is not the same as quantized sinq persistence." + "reason": "Different generation bugs with different triggers: compiled forward inconsistency vs flash-attention misgeneration." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Legacy BEiT field loading and Bamba compile/SDPA are different code paths with no common fix." + "reason": "Different model-loading vs compile/attention problems; no shared code-path bug." }, { - "left": "issue:44464", - "right": "issue:44977", + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both concern save/load behavior, but one is tied-weight serialization varying by device and the other is sinq quantization breaking after reload." + }, + { + "left": "issue:43782", + "right": "issue:46693", "accept": false, - "reason": "Both are generation regressions, but one is compiled-forward chunking and the other is FA2 behavior on Qwen3.5." + "reason": "Different root causes: Qwen3VL `from_pretrained(weight_only=True)` error vs a flash-attention long-context performance regression." }, { "left": "issue:42898", "right": "issue:46491", "accept": false, - "reason": "General v5 tokenization-space behavior change is not the same as the CodeLlama leading-space regression." + "reason": "Both are tokenization regressions, but one is a cleanup-space behavior change and the other is a CodeLlama leading-space round-trip bug." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Both are training-related, but one is a NoneType broadcast error and the other is a Qwen2-VL multi-GPU issue." - }, - { - "left": "issue:43782", - "right": "issue:46693", - "accept": false, - "reason": "Weight_only loading for Qwen3VL is unrelated to the FA2 long-context performance regression." + "reason": "Both involve training, but one is a `_gpu_broadcast_one` NoneType error and the other is Qwen2-VL multi-GPU training behavior." }, { "left": "issue:45405", "right": "issue:47254", "accept": false, - "reason": "Dependency version pinning and a PEFT/DeepSpeed checkpointing error are different problems." - }, - { - "left": "issue:43504", - "right": "issue:43720", - "accept": false, - "reason": "Different model families and failures: BEiT legacy field loading vs BitNet packed-weight loading." + "reason": "No overlap: an unreleased PEFT version bump vs a checkpointing error with PEFT + DeepSpeed ZeRO-3." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch and DiaConfig token-ID loss are unrelated." + "reason": "Different bugs: CPU offload device mismatch in bark-small vs DiaConfig losing custom token IDs after save/load." }, { "left": "issue:43720", - "right": "issue:43927", + "right": "issue:44361", "accept": false, - "reason": "BitNet packed-weight handling and DiaConfig save/load field loss do not share the same fix." + "reason": "Different areas and symptoms: BitNet packed-weight handling during accelerate loading vs MLukeTokenizer task-time AttributeError." }, { "left": "issue:43643", "right": "issue:44561", "accept": false, - "reason": "Both involve trust_remote_code, but one is missing config fields and the other is a removed helper breaking remote models." + "reason": "Related only by `trust_remote_code`, but one is missing fields from `AutoConfig.from_pretrained` and the other is a v5 removal breaking remote-code models." }, { - "left": "issue:43278", - "right": "issue:43688", + "left": "issue:43688", + "right": "issue:45440", "accept": false, - "reason": "Embedding dtype drift and auxiliary-loss normalization are unrelated model bugs." + "reason": "Both are MoE-related, but one is auxiliary-loss normalization and the other is a DeepSeekV3MoE implementation divergence." }, { - "left": "issue:43688", - "right": "issue:45440", + "left": "issue:43278", + "right": "issue:43688", "accept": false, - "reason": "Both touch MoE models, but the bugs are different: loss normalization vs native/remote implementation divergence." + "reason": "Different bugs: dtype mismatch between train/eval for embeddings vs auxiliary-loss normalization in MoE models." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Legacy preset loading for BEiT and OmDet-Turbo processor resolution are unrelated." + "reason": "Different vision-model bugs: legacy-field loading failure in BEiT vs OmDet-Turbo processor/model image-size mismatch." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Both mention SigLIP2, but one is a docs/example issue with model/processor mismatch and quantization failure, while the other is an implementation discrepancy vs JAX." }, { "left": "issue:44625", "right": "issue:47276", "accept": false, - "reason": "Qwen3.5 num_labels propagation and a misleading synced_gpus warning are different issues." + "reason": "Completely unrelated: Qwen3.5 `num_labels` propagation vs a misleading `synced_gpus` warning in continuous batching." }, { "left": "issue:43720", "right": "issue:44361", "accept": false, - "reason": "BitNet accelerated loading and MLukeTokenizer task-time AttributeError are unrelated." - }, - { - "left": "issue:44734", - "right": "issue:46612", - "accept": false, - "reason": "Serve KV-cache continuation indexing and beam-search cache reorder skipping are different generation paths." - }, - { - "left": "issue:43577", - "right": "issue:44554", - "accept": false, - "reason": "Blip2 dtype loading and MPS attention correctness are unrelated." + "reason": "Same as above: unrelated BitNet loading issue vs MLukeTokenizer task error; not mergeable as one bug fix." }, { - "left": "issue:43525", - "right": "issue:43572", + "left": "issue:43334", + "right": "issue:45759", "accept": false, - "reason": "Both are missing config token fields, but they affect different models and different attributes." + "reason": "Both are Qwen multimodal config issues, but one is a missing `pad_token_id` attribute and the other is `from_config` failing to unwrap `text_config`; different fixes and code paths." } ] }, @@ -8895,20 +8972,19 @@ "cluster_id": "cluster-10000-544", "nodes": [ "issue:28218", - "issue:29925", "issue:33357", "issue:39692", "issue:42796", "issue:43257", - "issue:43278", "issue:43295", "issue:43334", "issue:43493", + "issue:43525", "issue:43531", "issue:43572", + "issue:43577", "issue:43611", "issue:43646", - "issue:43701", "issue:43825", "issue:43927", "issue:43994", @@ -8922,13 +8998,13 @@ "issue:44554", "issue:44568", "issue:44704", + "issue:44734", "issue:44843", "issue:44964", "issue:44977", "issue:45290", "issue:45405", "issue:45491", - "issue:45759", "issue:45812", "issue:45823", "issue:46002", @@ -8937,6 +9013,7 @@ "issue:46209", "issue:46365", "issue:46489", + "issue:46612", "issue:46846", "issue:47332", "issue:47404", @@ -8944,231 +9021,231 @@ "issue:47448" ], "soft_pairs": [ - "issue:44964|issue:45405", + "issue:44734|issue:46612", + "issue:43577|issue:44554", "issue:45823|issue:46050", + "issue:43525|issue:43572", + "issue:44964|issue:45405", "issue:44514|issue:45290", + "issue:43334|issue:43572", + "issue:28218|issue:33357", "issue:44484|issue:44704", - "issue:43334|issue:45759", "issue:43611|issue:46846", - "issue:28218|issue:33357", "issue:39692|issue:43295", - "issue:43334|issue:43572", "issue:44977|issue:46209", + "issue:43994|issue:46192", "issue:44488|issue:44843", "issue:45405|issue:47332", - "issue:43994|issue:46192", - "issue:43257|issue:47404", "issue:42796|issue:44977", + "issue:43257|issue:47404", "issue:43493|issue:43825", - "issue:45491|issue:46002", - "issue:43927|issue:44361", "issue:43531|issue:47436", - "issue:44554|issue:46365", - "issue:43646|issue:44466", "issue:43572|issue:44291", - "issue:44568|issue:45812", - "issue:44568|issue:46489", + "issue:43927|issue:44361", + "issue:43646|issue:44466", + "issue:45491|issue:46002", "issue:46050|issue:47448", - "issue:43278|issue:43701", + "issue:44568|issue:45812", "issue:44493|issue:47448", - "issue:28218|issue:29925" + "issue:44554|issue:46365", + "issue:44568|issue:46489" ], - "prepared_review_unit_hash": "b243b2d77fd5c78fa19ccdbb725d2cacd740fa4d983da1bd097da4b46b0fa83e", + "prepared_review_unit_hash": "ad7ee392d670a5b0e4f047eb086922ec209a47952eb55f5a23b0dc015fc9e964", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36266, - "estimated_input_tokens": 9067, - "estimated_eval_tokens": 18390 + "serialized_chars": 36389, + "estimated_input_tokens": 9098, + "estimated_eval_tokens": 18452 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is not a duplicate set: it mixes many unrelated Transformer issues across tokenizers, multimodal processors, model loading, config regressions, and attention/offloading bugs. None of the soft pairs look like the same underlying bug or change, so they should not be merged as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "The candidate pairs are mostly semantically similar at a very broad subsystem level, but they do not describe the same concrete bug/change. I would keep them separate; this cluster looks like accidental topical similarity rather than true duplicates.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45812 is the strongest representative because it is a clear, high-signal v4\u2192v5 tokenizer regression with multiple inbound references and a concrete, reproducible failure mode.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:45812 is the most broadly relevant and well-documented issue in the set, with the clearest regression framing and the most external attention.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44964", - "right": "issue:45405", + "left": "issue:44734", + "right": "issue:46612", + "accept": false, + "reason": "Both involve generation cache behavior, but one is a KV-cache continuation crash in serve while the other is beam-search cache reorder being skipped for specific model families; different code paths and fixes." + }, + { + "left": "issue:43577", + "right": "issue:44554", "accept": false, - "reason": "One is a model-loading failure for Phi-4 multimodal; the other is a dependency version bump for PEFT. Different problem, different fix." + "reason": "One is a dtype propagation bug in BLIP2 loading, the other is an MPS attention correctness issue; unrelated mechanisms and symptoms." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Gemma4 device-mismatch during forward is unrelated to broken quantized save/load behavior. Same broad area, but different code paths." + "reason": "Device placement mismatch during Gemma4 forward and quantized save/load breakage are different problems with different failure points." }, { - "left": "issue:44514", - "right": "issue:45290", + "left": "issue:43525", + "right": "issue:43572", "accept": false, - "reason": "Both involve chat template/processor flows, but one is batching with padding=False and the other is assistant tool-call messages with no content. Distinct failures." + "reason": "Both are missing pad-token config attributes after v5, but they affect different model configs and different attribute names; not the same concrete bug." }, { - "left": "issue:44484", - "right": "issue:44704", + "left": "issue:44964", + "right": "issue:45405", "accept": false, - "reason": "Max shard size question is unrelated to AutoProcessor kwargs not reaching cached_file." + "reason": "A model-loading failure for Phi-4 multimodal is unrelated to a PEFT version bump that is too new for PyPI." }, { - "left": "issue:43334", - "right": "issue:45759", + "left": "issue:44514", + "right": "issue:45290", "accept": false, - "reason": "Qwen3-VL pad_token_id AttributeError is a config attribute bug; the other is AutoModelForCausalLM not unwrapping text_config for composite multimodal configs. Different issue." + "reason": "Both touch apply_chat_template, but one fails on batched padding=False input and the other on tool-call assistant messages with no content; distinct edge cases." }, { - "left": "issue:43611", - "right": "issue:46846", + "left": "issue:43334", + "right": "issue:43572", "accept": false, - "reason": "Both mention Qwen3.5/model loading behavior, but one is a base_model_prefix regression and the other is missing linear_attn TP plan entries causing OOM/shape errors. Not the same bug." + "reason": "Different model families and different missing config attributes; same broad regression class, but not one bug." }, { "left": "issue:28218", "right": "issue:33357", "accept": false, - "reason": "Tokenizer extra-space insertion is unrelated to a MacOS bus error on a community CLIP model." + "reason": "Tokenizer spacing after added tokens and a Mac bus error on a CLIP model are unrelated." }, { - "left": "issue:39692", - "right": "issue:43295", + "left": "issue:44484", + "right": "issue:44704", "accept": false, - "reason": "SigLIP2 example/doc errors are unrelated to a regression in processor.tokenizer access and image passing." + "reason": "One is a save_pretrained shard-size question, the other is AutoProcessor kwargs not reaching cached_file; no shared underlying defect." }, { - "left": "issue:43334", - "right": "issue:43572", + "left": "issue:43611", + "right": "issue:46846", + "accept": false, + "reason": "Base-model-prefix loading regression and missing tensor-parallel plan entries are separate model-loading/generation issues." + }, + { + "left": "issue:39692", + "right": "issue:43295", "accept": false, - "reason": "Both are config-attribute errors, but for different missing fields and different models (pad_token_id vs pad_token_idx)." + "reason": "Documentation example errors are not the same as a regression in processor.tokenizer access and image passing." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 flash-attention generation failure is unrelated to FP8 inference failing on B200." + "reason": "Flash-attention generation failure and FP8 inference failure on B200 are different hardware/model-path issues." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "Wrong SigLIP2 outputs and return_tensors being ignored are separate tokenizer/pipeline behaviors." }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "Model loading failure for cjvt/sleng-bert is unrelated to offline-mode breakage from an unconditional model_info() call." + "reason": "Loading a specific model and offline-mode breakage from unconditional model_info() calls are unrelated." }, { "left": "issue:45405", "right": "issue:47332", "accept": false, - "reason": "PEFT version bump and inability to save a disk-offloaded model are different problems with no shared code path." + "reason": "A dependency-version bump and inability to save an offloaded model are unrelated." }, { - "left": "issue:43994", - "right": "issue:46192", + "left": "issue:42796", + "right": "issue:44977", "accept": false, - "reason": "Granite tokenizer wrong IDs is a tokenizer logic regression; return_tensors being ignored with text_kwargs is a different API behavior issue." + "reason": "Tokenizer BPE marker handling and flash-attention generation failure are different bugs." }, { "left": "issue:43257", "right": "issue:47404", "accept": false, - "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed is unrelated to device_map loading OOM from transient contiguous() allocations." - }, - { - "left": "issue:42796", - "right": "issue:44977", - "accept": false, - "reason": "Ministral tokenizer BPE marker handling and Qwen3.5 flash-attention generation are separate bugs in different components." + "reason": "Qwen3 MoE weight conversion and device-map loading OOM from conversion transients are not the same change." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 HF/JAX discrepancy is unrelated to a pipeline error message incorrectly suggesting translation tasks in v5." - }, - { - "left": "issue:45491", - "right": "issue:46002", - "accept": false, - "reason": "Gemma3 NaNs from sliding-window/all-padding attention are unrelated to classification heads producing NaNs after checkout+init." + "reason": "SigLIP2 numerical discrepancy and a pipeline error-message regression are unrelated." }, { - "left": "issue:43927", - "right": "issue:44361", + "left": "issue:43531", + "right": "issue:47436", "accept": false, - "reason": "DiaConfig losing custom token IDs on save/load is unrelated to MLukeTokenizer AttributeError on tasks." + "reason": "Sliding-window behavior in Qwen3-MoE and a missing max_position_embeddings config attribute are separate issues." }, { - "left": "issue:43531", - "right": "issue:47436", + "left": "issue:43572", + "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window issue is not the same as Qwen2_5OmniConfig missing max_position_embeddings." + "reason": "Missing pad_token_idx after v5 and an unexpected _is_hf_initialized init argument are different migration regressions." }, { - "left": "issue:44554", - "right": "issue:46365", + "left": "issue:43927", + "right": "issue:44361", "accept": false, - "reason": "MPS attention correctness bug has nothing to do with importing Trainer conflicting with datasets." + "reason": "DiaConfig save/load token-ID loss and MLukeTokenizer task AttributeError are distinct tokenizer/config bugs." }, { "left": "issue:43646", "right": "issue:44466", "accept": false, - "reason": "Custom model initialization breakage and inconsistent lm_head.weight serialization are different regression classes." + "reason": "Custom model initialization breaks in v5, but inconsistent lm_head.weight serialization is a different issue." }, { - "left": "issue:43572", - "right": "issue:44291", + "left": "issue:45491", + "right": "issue:46002", "accept": false, - "reason": "Missing pad_token_idx after 5.0 update and unexpected _is_hf_initialized argument with init_empty_weights are separate config/init issues." + "reason": "Both mention NaNs, but the causes and affected components differ completely." }, { - "left": "issue:44568", - "right": "issue:45812", + "left": "issue:46050", + "right": "issue:47448", "accept": false, - "reason": "MDeBERTa special-token handling and Granite AutoTokenizer wrong IDs are both tokenizer bugs, but not the same model or failure mode." + "reason": "Quantized model save/load breakage and an unrecognized Gemma4 architecture are unrelated." }, { "left": "issue:44568", - "right": "issue:46489", + "right": "issue:45812", "accept": false, - "reason": "MDeBERTa BOS/EOS omission is unrelated to DeepSeek-Coder v1 incorrect output and a gap in a prior tokenizer fix." + "reason": "Both are tokenizer regressions, but one is special-token insertion for mdeberta and the other is wrong token IDs for Granite; not the same concrete bug." }, { - "left": "issue:46050", + "left": "issue:44493", "right": "issue:47448", "accept": false, - "reason": "Quantized model save/load breakage is unrelated to Gemma4 architecture not being recognized." - }, - { - "left": "issue:43278", - "right": "issue:43701", - "accept": false, - "reason": "BF16-to-FP32 embedding dtype change is unrelated to resume_from_checkpoint key mismatch." + "reason": "Unexpected position-id keys and unrecognized Gemma4 architecture are different model/config problems." }, { - "left": "issue:44493", - "right": "issue:47448", + "left": "issue:44554", + "right": "issue:46365", "accept": false, - "reason": "Unexpected position-id keys and Gemma4 arch recognition are distinct model/config loading problems." + "reason": "An attention correctness issue on MPS is unrelated to an import conflict involving datasets and Trainer." }, { - "left": "issue:28218", - "right": "issue:29925", + "left": "issue:44568", + "right": "issue:46489", "accept": false, - "reason": "Additional space after added token and SigLIP tokenizer not enforcing use_fast=True are different tokenizer issues with different symptoms and fixes." + "reason": "Different tokenizer regressions affecting different models and likely different code paths; too broad to merge." } ] }, @@ -9194,6 +9271,7 @@ "nodes": [ "issue:15354", "issue:28218", + "issue:29925", "issue:34567", "issue:37428", "issue:38175", @@ -9223,7 +9301,6 @@ "issue:44964", "issue:45072", "issue:45084", - "issue:45200", "issue:45229", "issue:45245", "issue:45357", @@ -9242,252 +9319,252 @@ "issue:47436" ], "soft_pairs": [ + "issue:43278|issue:43701", + "issue:28218|issue:29925", "issue:43873|issue:47404", - "issue:43278|issue:43844", - "issue:43493|issue:43844", + "issue:42490|issue:43696", "issue:43504|issue:43606", - "issue:43824|issue:44589", + "issue:43493|issue:43844", "issue:46821|issue:46832", + "issue:43824|issue:44589", "issue:44589|issue:45520", + "issue:43278|issue:43844", "issue:46209|issue:46710", - "issue:42490|issue:43696", - "issue:43504|issue:43653", + "issue:43994|issue:44493", "issue:43606|issue:44112", + "issue:43504|issue:43653", "issue:44493|issue:46285", - "issue:43994|issue:44493", "issue:45072|issue:46832", "issue:43873|issue:46050", "issue:44464|issue:45584", + "issue:44423|issue:44734", "issue:15354|issue:38175", "issue:43572|issue:47436", - "issue:45245|issue:46710", "issue:47013|issue:47254", - "issue:44849|issue:45357", "issue:45084|issue:45520", - "issue:44423|issue:44734", + "issue:45245|issue:46710", "issue:42947|issue:43381", + "issue:44849|issue:45357", "issue:45229|issue:45823", - "issue:43844|issue:44964", "issue:28218|issue:34567", + "issue:43844|issue:44964", "issue:44849|issue:46209", - "issue:45200|issue:45229", - "issue:37428|issue:43957", - "issue:43493|issue:43701" + "issue:37428|issue:43957" ], - "prepared_review_unit_hash": "4b143798ddc1f8741a883777d42684933ec23f09a3d07797dbf64054c9783792", + "prepared_review_unit_hash": "ba23d0a69162676b807f4e884d5ff0b3fa4712d171aba629bcb766733995ed23", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 37137, - "estimated_input_tokens": 9285, - "estimated_eval_tokens": 18826 + "serialized_chars": 37082, + "estimated_input_tokens": 9271, + "estimated_eval_tokens": 18798 }, "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; the soft links are too weak to treat any pair as duplicates. No PRs are present.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue fits well because the items span unrelated bugs across loading, tokenization, training, inference, and CI.", + "summary": "All items are issues, not PRs. The cluster is a loose mix of unrelated Transformers bugs; none of the soft pairs look like the same underlying problem, and issue 43873 is the strongest single anchor by activity and breadth.", + "confidence": 0.93, + "canonical_issue_reason": "No true duplicate cluster is evident; issue 43873 is the most active and broadly framed bug report, so it is the best anchor among these issues.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43957 is the broadest standalone report: it describes a concrete model-loading regression affecting multiple models under Transformers 5.x.", + "best_issue_reason": "Issue 43873 has the highest engagement and a clear reproducible bug description, making it the best overall representative of the set.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43873", - "right": "issue:47404", + "left": "issue:43278", + "right": "issue:43701", "accept": false, - "reason": "Both mention offloading/memory, but one is about quantization offload behavior and the other about transient OOMs during device-map loading; not the same bug." + "reason": "Different symptoms and code paths: eval dtype drift vs checkpoint resume key mismatch." }, { - "left": "issue:43278", - "right": "issue:43844", + "left": "issue:28218", + "right": "issue:29925", "accept": false, - "reason": "BF16-to-FP32 dtype drift in eval is unrelated to the ZeRO-3 gradient growth problem." + "reason": "Tokenizer spacing after added token is unrelated to SigLIP fast-tokenizer enforcement." }, { - "left": "issue:43493", - "right": "issue:43844", + "left": "issue:43873", + "right": "issue:47404", + "accept": false, + "reason": "Both involve loading/offloading, but one is quantization offload behavior and the other is transient OOM during conversion." + }, + { + "left": "issue:42490", + "right": "issue:43696", "accept": false, - "reason": "SigLIP2 HF-vs-JAX discrepancy is a model fidelity issue, not the same as a DeepSpeed gradient anomaly." + "reason": "Import warning on trailing slash is unrelated to a GPT-oss CUDA OOM." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Legacy-field load failure in Beit and CPU offload device mismatch in Bark are unrelated code paths." + "reason": "Legacy-field model loading failure is unrelated to CPU offload device mismatch." }, { - "left": "issue:43824", - "right": "issue:44589", + "left": "issue:43493", + "right": "issue:43844", "accept": false, - "reason": "Importing a missing Qwen VL class is unrelated to the Float8 storage lookup failure." + "reason": "SigLIP2 HF/JAX discrepancy is a model correctness issue, not the ZeRO-3 gradient blowup described here." }, { "left": "issue:46821", "right": "issue:46832", "accept": false, - "reason": "Different models and different failures: TimesFM attribute error vs RT-DETR feature-level mismatch." + "reason": "Different model bugs: TimesFM window_size attribute error vs RT-DETR feature-level mismatch." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing a missing Qwen class is not the same as a Float8 storage lookup failure." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage resolution and flash_attn import-key handling are distinct low-level import/runtime issues." + "reason": "Float8 storage resolution and flash_attn import-key handling are separate import/runtime issues." }, { - "left": "issue:46209", - "right": "issue:46710", + "left": "issue:43278", + "right": "issue:43844", "accept": false, - "reason": "FP8 inference on B200 hardware and an accuracy regression after upgrade are not the same concrete bug." + "reason": "Eval dtype change and gradient explosion under ZeRO-3 are unrelated failures." }, { - "left": "issue:42490", - "right": "issue:43696", + "left": "issue:46209", + "right": "issue:46710", "accept": false, - "reason": "Path-trailing-slash import warning is unrelated to a CUDA OOM on GPT-oss-20b." + "reason": "FP8 inference OOM on B200 is not the same as an accuracy regression after upgrade." }, { - "left": "issue:43504", - "right": "issue:43653", + "left": "issue:43994", + "right": "issue:44493", "accept": false, - "reason": "Beit pretrained-load legacy-field handling is unrelated to BigBirdTokenizer special-token registration." + "reason": "SigLIP2 nonsense outputs and unexpected position-id keys are distinct model bugs." }, { "left": "issue:43606", "right": "issue:44112", "accept": false, - "reason": "Both involve device issues, but the Bark CPU-offload mismatch and GraniteSpeech CI stale override are different failures." + "reason": "A real device-mismatch bug is not the same as a stale CI test." }, { - "left": "issue:44493", - "right": "issue:46285", + "left": "issue:43504", + "right": "issue:43653", "accept": false, - "reason": "Unexpected position-id keys across models is not the same bug as the CLIP text-model structure change." + "reason": "Pretrained-load legacy field error is unrelated to BigBirdTokenizer mask-token registration." }, { - "left": "issue:43994", - "right": "issue:44493", + "left": "issue:44493", + "right": "issue:46285", "accept": false, - "reason": "SigLIP2 nonsensical outputs and unexpected key warnings have different symptoms and likely different causes." + "reason": "Unexpected position-id keys and CLIP text-model structural inconsistency are different regressions." }, { "left": "issue:45072", "right": "issue:46832", "accept": false, - "reason": "BF16 dtype mismatches in inference are unrelated to RT-DETR's feature-level indexing error." + "reason": "Bfloat16 dtype mismatch in inference is unrelated to RT-DETR forward shape/length error." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Quantization offloading problems and save/load breakage after sinq quantization are separate issues." + "reason": "Quantized offloading and save/load breakage for sinq are different quantization bugs." }, { "left": "issue:44464", "right": "issue:45584", "accept": false, - "reason": "Chunked generation with compiled forward and Whisper empty-transcription handling are distinct generation bugs." + "reason": "Compiled-forward chunked generation inconsistency is unrelated to Whisper empty-transcription failure." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing are different serve paths." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "TorchScript GeneratorExp export failure is unrelated to SigLIP2 zero-probability outputs." + "reason": "TorchScript GeneratorExp limitation is unrelated to SigLIP2 zero-probability outputs." }, { "left": "issue:43572", "right": "issue:47436", "accept": false, - "reason": "Both are missing config fields, but for different models and different attributes; not a duplicate." - }, - { - "left": "issue:45245", - "right": "issue:46710", - "accept": false, - "reason": "Category-count runtime error and an accuracy regression are unrelated." + "reason": "Missing pad_token_idx and missing max_position_embeddings are different config regressions." }, { "left": "issue:47013", "right": "issue:47254", "accept": false, - "reason": "xLSTM unpacking crash and PEFT+DeepSpeed checkpoint errors are different training-time failures." - }, - { - "left": "issue:44849", - "right": "issue:45357", - "accept": false, - "reason": "Qwen3.5 output_hidden_states behavior and incorrect save_pretrained visual keys are separate bugs." + "reason": "xLSTM train-mode unpacking crash is unrelated to PEFT+ZeRO-3 checkpoint error." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Template-node compilation errors are unrelated to flash_attn import key errors." + "reason": "Template-node compilation failure is unrelated to flash_attn import lookup on Python 3.13." }, { - "left": "issue:44423", - "right": "issue:44734", + "left": "issue:45245", + "right": "issue:46710", "accept": false, - "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing are different serve-path bugs." + "reason": "Category-count limit error is unrelated to the DeepSeek accuracy regression." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Gradient checkpointing with PEFT LoRA effectiveness and eval-mode prohibition are related only broadly, not the same defect." + "reason": "Gradient-checkpointing ineffectiveness in PEFT LoRA is not the same as being unsupported in eval mode." }, { - "left": "issue:45229", - "right": "issue:45823", + "left": "issue:44849", + "right": "issue:45357", "accept": false, - "reason": "Gemma4 OOM during multi-GPU inference and Gemma4 PLE device mismatch are different runtime issues." + "reason": "Qwen3.5 hidden-states bug and incorrect visual-encoder save keys are different regressions." }, { - "left": "issue:43844", - "right": "issue:44964", + "left": "issue:45229", + "right": "issue:45823", "accept": false, - "reason": "ZeRO-3 gradient instability is unrelated to loading Phi-4-multimodal-instruct." + "reason": "Multi-GPU CUDA OOM and device mismatch with device_map='auto' are separate Gemma4 issues." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Tokenizer spacing after added tokens and TrainerState token counting are unrelated subsystems." + "reason": "Tokenizer added-token spacing is unrelated to TrainerState token-count tracking." }, { - "left": "issue:44849", - "right": "issue:46209", + "left": "issue:43844", + "right": "issue:44964", "accept": false, - "reason": "Qwen3.5 hidden-state behavior and FP8 inference on B200 are unrelated problems." + "reason": "ZeRO-3 gradient anomaly is unrelated to Phi-4 multimodal loading failure." }, { - "left": "issue:45200", - "right": "issue:45229", + "left": "issue:44849", + "right": "issue:46209", "accept": false, - "reason": "Gemma4 text-only mm_token_type_ids defaults and Gemma4 multi-GPU OOM are different issues." + "reason": "Qwen3.5 hidden-states bug is unrelated to FP8 inference failure on B200." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Flash-attention import failure and meta-device model loading breakage are unrelated." - }, - { - "left": "issue:43493", - "right": "issue:43701", - "accept": false, - "reason": "SigLIP2 implementation mismatch and resume-from-checkpoint key mismatch are unrelated." + "reason": "Flash-attention import failure and meta-device model loading breakage are different infrastructure bugs." } ] }, @@ -9511,24 +9588,24 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ - "issue:27957", "issue:28370", "issue:29942", "issue:30333", "issue:34689", "issue:37428", "issue:41720", - "issue:42915", "issue:43065", "issue:43295", "issue:43334", "issue:43404", "issue:43450", + "issue:43493", "issue:43525", "issue:43540", "issue:43550", "issue:43572", "issue:43575", + "issue:43701", "issue:43824", "issue:43844", "issue:43873", @@ -9542,14 +9619,13 @@ "issue:45072", "issue:45198", "issue:45200", + "issue:45229", "issue:45406", "issue:45446", - "issue:45520", "issue:45588", "issue:45759", "issue:45823", "issue:46121", - "issue:46178", "issue:46192", "issue:46369", "issue:46701", @@ -9561,238 +9637,231 @@ "issue:47404" ], "soft_pairs": [ + "issue:45200|issue:45229", "issue:41720|issue:46710", + "issue:43493|issue:43701", + "issue:47332|issue:47404", "issue:43937|issue:46701", "issue:44849|issue:47145", - "issue:47332|issue:47404", + "issue:43295|issue:43450", "issue:44977|issue:46701", "issue:43065|issue:46369", - "issue:43295|issue:43450", - "issue:44554|issue:45446", "issue:43404|issue:46710", - "issue:45072|issue:45198", + "issue:44554|issue:45446", "issue:45588|issue:45759", - "issue:44869|issue:46752", + "issue:45072|issue:45198", "issue:43295|issue:43873", + "issue:44869|issue:46752", "issue:43572|issue:46821", "issue:29942|issue:30333", "issue:43540|issue:45200", + "issue:45823|issue:46192", "issue:28370|issue:34689", "issue:45406|issue:46121", "issue:43295|issue:44610", - "issue:45823|issue:46192", "issue:44186|issue:46192", + "issue:43334|issue:43525", "issue:43450|issue:43575", "issue:37428|issue:43824", - "issue:43550|issue:43844", "issue:45200|issue:45823", - "issue:45520|issue:46178", - "issue:43334|issue:43525", - "issue:42915|issue:43824", - "issue:27957|issue:28370" + "issue:43550|issue:43844" ], - "prepared_review_unit_hash": "8d6e7052d980c0fbafd75c3d5c2efc07645a85200f50aa439a7c6bd0ad6f9c42", + "prepared_review_unit_hash": "f31e91eaba5ad0a2f10fb6933f76b3c36a6dac4fd555563da707417c51650caf", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36402, - "estimated_input_tokens": 9101, - "estimated_eval_tokens": 18458 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35518, + "estimated_input_tokens": 8880, + "estimated_eval_tokens": 18016 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose similarity group of unrelated Transformers issues across model loading, processor/tokenizer handling, flash-attention, offloading, and config regressions. None of the soft pairs look like true duplicates, so I would not merge any of them.", - "confidence": 0.44, - "canonical_issue_reason": "Issue 43295 is the broadest and most central regression here: it describes a framework-level processor/tokenizer break that is closest to several other multimodal input-handling issues in the set.", + "summary": "This looks like a loose topical cluster of mostly unrelated Transformers issues (model loading, multimodal processors, offloading, flash-attn, configs). I do not see any soft pair that is clearly the same underlying bug/change, so no canonical duplicate stands out.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "43295 is the best anchor issue for the cluster because it is a clear cross-cutting regression rather than a narrow model-specific symptom.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:41720", - "right": "issue:46710", + "left": "issue:45200", + "right": "issue:45229", "accept": false, - "reason": "Both mention Qwen models, but one is a cudaErrorAssert during auto device mapping and the other is an accuracy regression after an upgrade; different bugs and code paths." + "reason": "Gemma4 tokenizer/input-default bug vs multi-GPU inference OOM; different failure modes and code paths." }, { - "left": "issue:43937", - "right": "issue:46701", + "left": "issue:41720", + "right": "issue:46710", "accept": false, - "reason": "GLM-5 invalid GenerationConfig and DiffusionGemma pipeline failure are unrelated model families and unrelated failure modes." + "reason": "Qwen3 auto device-map CUDA assert vs DeepSeek accuracy regression after upgrade; unrelated models and symptoms." }, { - "left": "issue:44849", - "right": "issue:47145", + "left": "issue:43493", + "right": "issue:43701", "accept": false, - "reason": "Qwen3.5 output_hidden_states bug and FP8 frozen-checkpoint weight_scale loading bug are different mechanisms with no shared concrete code path." + "reason": "SigLIP2 implementation discrepancy vs resume-from-checkpoint key mismatch; no shared concrete bug." }, { "left": "issue:47332", "right": "issue:47404", "accept": false, - "reason": "Both involve offloading/loading, but one is a save failure and the other is an OOM during device_map conversion; not the same underlying bug." + "reason": "Both involve loading/offloading, but one is save-time failure for a specific model and the other is load-time OOM from conversion transients; not the same bug." }, { - "left": "issue:44977", + "left": "issue:43937", "right": "issue:46701", "accept": false, - "reason": "Flash-attention generation issues in Qwen3.5 and a DiffusionGemma image-text pipeline failure are not the same problem." + "reason": "Invalid GenerationConfig vs image-text-to-text pipeline failure; different subsystems and root causes." }, { - "left": "issue:43065", - "right": "issue:46369", + "left": "issue:44849", + "right": "issue:47145", "accept": false, - "reason": "Both are Sam3-related, but one is a dummy Conv2d architectural issue and the other is a memory overflow when calling the model." + "reason": "Qwen3.5 hidden_states regression vs FP8 weight_scale not applied on load; distinct correctness issues." }, { "left": "issue:43295", "right": "issue:43450", "accept": false, - "reason": "Processor/tokenizer regression and batched video shape mismatch are different processor bugs with different affected paths." + "reason": "Processor.tokenizer regression vs batched video input shape bug; separate processor problems." }, { - "left": "issue:44554", - "right": "issue:45446", + "left": "issue:44977", + "right": "issue:46701", "accept": false, - "reason": "MPS attention correctness and a PyTorch version check for AuxRequest import are unrelated." + "reason": "Flash-attention generation bug vs DiffusionGemma pipeline not working; unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Sam3PixelDecoder dummy layer vs sam3-litetext memory overflow; different implementation issues." }, { "left": "issue:43404", "right": "issue:46710", "accept": false, - "reason": "Mistral3 lm_head weight tying and DeepSeek accuracy regression are different models and different failure classes." + "reason": "lm_head weight tying bug vs accuracy regression; not the same code-path." }, { - "left": "issue:45072", - "right": "issue:45198", + "left": "issue:44554", + "right": "issue:45446", "accept": false, - "reason": "Bfloat16 dtype mismatches in CI and Wav2Vec2 save_pretrained/tokenization failures do not share a concrete bug." + "reason": "MPS attention correctness bug vs flex_attention import version check; different layers entirely." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "One is a flash_attention sink-less crash on s_aux=None; the other is AutoModelForCausalLM.from_config not unwrapping text_config. Similar area, but not the same bug." + "reason": "flash_attention.py s_aux=None crash vs AutoModelForCausalLM.from_config text_config unwrap issue; unrelated." }, { - "left": "issue:44869", - "right": "issue:46752", + "left": "issue:45072", + "right": "issue:45198", "accept": false, - "reason": "Whisper timestamp decoding on a trailing replacement character and apply_chat_template on an empty list are separate tokenizer edge cases." + "reason": "bfloat16 dtype mismatch bugs vs Wav2Vec2 save_pretrained/tokenization failure; not a duplicate." }, { "left": "issue:43295", "right": "issue:43873", "accept": false, - "reason": "Processor/tokenizer regression and quantization offloading not working are different subsystems and failure modes." + "reason": "Processor.tokenizer regression vs quantized offloading behavior; only broad subsystem overlap." + }, + { + "left": "issue:44869", + "right": "issue:46752", + "accept": false, + "reason": "Whisper timestamp decoding crash vs empty chat-template IndexError; different text-processing paths." }, { "left": "issue:43572", "right": "issue:46821", "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig and TimesFM window_size AttributeError are separate config/attribute bugs." + "reason": "Missing pad_token_idx vs missing window_size attribute; different config fields and models." }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "Failing Flash Attention 2 tests and an MLFlow reporting issue are unrelated." + "reason": "Flash Attention 2 test failures vs MLflow reporting on failed training jobs; unrelated." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Qwen3OmniMoe video processing ValueError and Gemma 4 mm_token_type_ids defaults are both multimodal, but they are not the same underlying defect." + "reason": "Qwen3OmniMoe video-processing ValueError vs Gemma4 token-type-id defaulting; separate bugs." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device_map auto placement mismatch vs return_tensors being ignored; different behavior and fixes." }, { "left": "issue:28370", "right": "issue:34689", "accept": false, - "reason": "Custom SentencePiece vocab_file handling and Llama 3.2 Vision model loading regression are different model-loading bugs." + "reason": "Missing vocab_file for custom SentencePiece models vs Llama 3.2 Vision load breakage; different model-loading problems." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing _tokenizer and rope-param list TypeError are unrelated code paths." + "reason": "Gemma4Processor missing _tokenizer vs rope validation list-type error; no shared underlying bug." }, { "left": "issue:43295", "right": "issue:44610", "accept": false, - "reason": "Both involve processors, but one is a tokenizer/processor regression and the other is an image-size mismatch for OmDet-Turbo." + "reason": "processor.tokenizer regression vs OmDet processor image-size mismatch; different processor code paths." }, { - "left": "issue:45823", + "left": "issue:44186", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device mismatch with device_map='auto' and return_tensors being ignored are different bugs." + "reason": "LayoutLMv2Tokenizer NER padding/truncation crash vs return_tensors being ignored; unrelated tokenization issues." }, { - "left": "issue:44186", - "right": "issue:46192", + "left": "issue:43334", + "right": "issue:43525", "accept": false, - "reason": "LayoutLMv2Tokenizer crashing on NER/batched padding and return_tensors being ignored are separate tokenizer behaviors." + "reason": "Both mention missing pad_token_id, but they affect different model families/config classes and likely need separate fixes." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Video processor shape issues and Qwen2-57B-A14B load OOM are unrelated." + "reason": "Batched video processor shape bug vs tensor-parallel OOM on Qwen2 loading; no common bug." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different modules/models." - }, - { - "left": "issue:43550", - "right": "issue:43844", - "accept": false, - "reason": "torch.compile/SDPA failure in Bamba-9B-v2 and a ZeRO-3 gradient explosion issue are different training/runtime bugs." + "reason": "Missing flash-attention support import vs missing Qwen2.5-VL class import; unrelated import errors." }, { "left": "issue:45200", "right": "issue:45823", "accept": false, - "reason": "Same Gemma 4 family, but one is a missing default input field and the other is a device placement mismatch; not the same fix." - }, - { - "left": "issue:45520", - "right": "issue:46178", - "accept": false, - "reason": "Python 3.13 flash_attn import KeyError and tensorflow_text backend mapping failure are unrelated import/configuration problems." - }, - { - "left": "issue:43334", - "right": "issue:43525", - "accept": false, - "reason": "Both mention missing pad_token_id, but they affect different model configs and different loading paths; too broad to be one duplicate." - }, - { - "left": "issue:42915", - "right": "issue:43824", - "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure and a Qwen2.5-VL import error are unrelated." + "reason": "Gemma4 mm_token_type_ids defaulting issue vs Gemma4 PLE device mismatch; separate defects." }, { - "left": "issue:27957", - "right": "issue:28370", + "left": "issue:43550", + "right": "issue:43844", "accept": false, - "reason": "XLMRoberta Flash Attention 2 and custom SentencePiece vocab_file handling are completely different issues." + "reason": "torch.compile SDPA failure in Bamba vs DeepSpeed ZeRO-3 gradient anomaly; different training/runtime bugs." } ] }, @@ -9816,12 +9885,15 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ + "issue:27957", "issue:28218", + "issue:28370", "issue:29870", "issue:29942", "issue:35141", "issue:39692", "issue:42490", + "issue:42915", "issue:43262", "issue:43278", "issue:43295", @@ -9842,18 +9914,15 @@ "issue:44466", "issue:44521", "issue:44561", - "issue:44610", "issue:44737", - "issue:44857", - "issue:44898", "issue:45127", "issue:45198", "issue:45200", "issue:45290", "issue:45325", "issue:45381", - "issue:45412", "issue:45520", + "issue:46178", "issue:46192", "issue:46209", "issue:46234", @@ -9865,217 +9934,224 @@ "issue:47448" ], "soft_pairs": [ + "issue:45520|issue:46178", + "issue:42915|issue:43824", + "issue:39692|issue:46192", + "issue:27957|issue:28370", "issue:29870|issue:29942", "issue:42490|issue:43493", - "issue:39692|issue:46192", "issue:43295|issue:43493", "issue:43824|issue:43975", "issue:43278|issue:43856", - "issue:44521|issue:45290", "issue:46192|issue:47448", - "issue:43262|issue:43450", "issue:43975|issue:45520", + "issue:43262|issue:43450", + "issue:44521|issue:45290", "issue:28218|issue:35141", - "issue:43653|issue:44112", "issue:43950|issue:46285", + "issue:43653|issue:44112", "issue:44737|issue:46894", "issue:43452|issue:46512", - "issue:44561|issue:46234", "issue:45200|issue:45325", - "issue:44466|issue:45127", + "issue:44561|issue:46234", "issue:43819|issue:44112", + "issue:44466|issue:45127", "issue:44442|issue:45198", "issue:43335|issue:43550", "issue:45381|issue:46209", - "issue:43611|issue:47405", - "issue:44898|issue:45412", "issue:45200|issue:46899", - "issue:44610|issue:44857" + "issue:43611|issue:47405" ], - "prepared_review_unit_hash": "d04466aa362f3f2c302d41f97a60b8a9e79d7f58dd8fa51a2c75858f8ce2cd44", + "prepared_review_unit_hash": "8a655a4bfe9a60298717d729785264c805f3cb70360ff07d06af51345b2fde5e", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 26, - "serialized_chars": 35421, - "estimated_input_tokens": 8856, - "estimated_eval_tokens": 17968 + "soft_pair_count": 27, + "serialized_chars": 35664, + "estimated_input_tokens": 8916, + "estimated_eval_tokens": 18088 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of mostly unrelated Transformers issues across tokenizers, processors, model loading, and modality-specific bugs. None of the soft pairs look like the same underlying defect, so I would not merge any of them as duplicates.", - "confidence": 0.94, - "canonical_issue_reason": "No clear canonical issue: the cluster is heterogeneous and the soft links reflect topical similarity, not true duplicates.", + "summary": "This cluster is not a duplicate set: it mixes many unrelated bug reports across tokenization, model loading, multimodal processing, quantization, CI, and architecture-specific regressions. None of the soft-similarity edges look like the same underlying bug/change.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out because the items are heterogeneous and the proposed links are only superficial; they do not share a concrete code path or failure mode.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the broadest user-facing representative of the multimodal/chat-template regressions, but it is not a duplicate anchor for the rest of the items.", + "best_issue_reason": "If a representative issue is needed, issue 43493 is the strongest standalone report: it is specific, reproducible, and describes a clear model-vs-JAX discrepancy.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:29870", - "right": "issue:29942", + "left": "issue:45520", + "right": "issue:46178", "accept": false, - "reason": "Both involve generation/testing failures, but one is a Bart decoder attention mask bug and the other is Flash Attention 2 test failures; different code paths and fixes." + "reason": "Different subsystems and failures: flash_attn import key error vs backend registry validation for tensorflow_text." }, { - "left": "issue:42490", - "right": "issue:43493", + "left": "issue:42915", + "right": "issue:43824", "accept": false, - "reason": "Local-path import warning vs SigLIP2 implementation discrepancy; unrelated problems." + "reason": "Unrelated model-specific FP8 inference failure vs an import error for a missing Qwen2.5-VL class." }, { "left": "issue:39692", "right": "issue:46192", "accept": false, - "reason": "Docs/example errors and a return_tensors/text_kwargs API bug are different issues." + "reason": "Docs/example errors are not the same bug as return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:27957", + "right": "issue:28370", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to missing vocab_file on custom SentencePiece models." + }, + { + "left": "issue:29870", + "right": "issue:29942", + "accept": false, + "reason": "A Bart generate() UnboundLocalError is not the same as failing Flash Attention 2 tests." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Local-path import warning and SigLIP2 implementation discrepancy affect different loading/behavior paths." }, { "left": "issue:43295", "right": "issue:43493", "accept": false, - "reason": "Processor/tokenizer regression versus SigLIP2 model parity issue; not the same bug." + "reason": "Processor/tokenizer regression in custom model code is unrelated to the SigLIP2 JAX-vs-HF discrepancy." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "Qwen2.5-VL import failure and DeepSeek detokenization regression are distinct model-specific failures." + "reason": "ImportError for a vision model class and detokenization for DeepSeek are distinct bugs." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "Embedding dtype drift in evaluation and Qwen3 MoE memory usage are unrelated." + "reason": "A dtype regression during evaluate is not the same as Qwen3 MoE memory inefficiency during training." }, { - "left": "issue:44521", - "right": "issue:45290", + "left": "issue:46192", + "right": "issue:47448", "accept": false, - "reason": "Both touch chat templating, but one is assistant-mask generation for multimodal inputs and the other is a crash on tool-call messages with no content; not the same concrete defect." + "reason": "Ignored return_tensors behavior is unrelated to Gemma4 architecture recognition/loading." }, { - "left": "issue:46192", - "right": "issue:47448", + "left": "issue:43975", + "right": "issue:45520", "accept": false, - "reason": "A tokenizer-kwargs bug and a Gemma4 architecture recognition issue are unrelated." + "reason": "Detokenization output bugs and flash_attn import issues do not share a code path." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template sampling-rate default vs batched video processor shape bug; different modalities and code paths." + "reason": "Audio chat-template sampling-rate handling is unrelated to batched video processor output shape." }, { - "left": "issue:43975", - "right": "issue:45520", + "left": "issue:44521", + "right": "issue:45290", "accept": false, - "reason": "Incorrect detokenization and a Python 3.13 import_utils KeyError are unrelated." + "reason": "Both touch chat templates, but one is multimodal assistant masks and the other is tool-call tokenization crash; different bugs." }, { "left": "issue:28218", "right": "issue:35141", "accept": false, - "reason": "Extra spacing after an added token and output embedding reinitialization after resizing are different tokenizer/model-weight issues." + "reason": "Extra space after added token is a tokenizer formatting issue, not the same as embedding reinitialization on resize." }, { - "left": "issue:43653", - "right": "issue:44112", + "left": "issue:43950", + "right": "issue:46285", "accept": false, - "reason": "BigBirdTokenizer special-token decode bug versus an unrelated GraniteSpeech CI test failure." + "reason": "Non-persistent buffer corruption in from_pretrained is unrelated to CLIP text model structure divergence." }, { - "left": "issue:43950", - "right": "issue:46285", + "left": "issue:43653", + "right": "issue:44112", "accept": false, - "reason": "Non-persistent buffer corruption in from_pretrained is a different bug from CLIP text model structural inconsistency." + "reason": "BigBirdTokenizer special-token registration and GraniteSpeech CI device override are unrelated issues." }, { "left": "issue:44737", "right": "issue:46894", "accept": false, - "reason": "XLNet CPU device placement bug and CohereASR training-loss shift bug do not share the same code path." + "reason": "CPU arange device placement in XLNet is not the same as the CohereASR loss double-shift bug." }, { "left": "issue:43452", "right": "issue:46512", "accept": false, - "reason": "gguf_file loading failure and stale config.dtype after from_config are separate loading/configuration issues." - }, - { - "left": "issue:44561", - "right": "issue:46234", - "accept": false, - "reason": "Both are missing import_utils helpers, but they are different flags with different breakages; not a duplicate fix." + "reason": "gguf_file loading failures are unrelated to from_config leaving config.dtype stale." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Gemma4 token-type defaults and Qwen2.5-VL rope/position-id scaling are different multimodal bugs." + "reason": "Gemma4 mm_token_type_ids defaults and Qwen2.5-VL rope indexing are different multimodal bugs." }, { - "left": "issue:44466", - "right": "issue:45127", + "left": "issue:44561", + "right": "issue:46234", "accept": false, - "reason": "Serialization of lm_head.weight by device and LoRA merge collapse with extended vocab are related only at a high level, not the same defect." + "reason": "Removing is_torch_fx_available and missing is_falcon_mamba_ssm_available are separate import_utils regressions." }, { "left": "issue:43819", "right": "issue:44112", "accept": false, - "reason": "DAC latent mismatch and GraniteSpeech CI flake are unrelated." + "reason": "DAC forward/from_latents mismatch is unrelated to a stale CI test on GraniteSpeech." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Serialization inconsistency for lm_head.weight is not the same as model collapse after LoRA merge with extended vocabulary." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "FastSpeech2ConformerTokenizer loading failure and Wav2Vec2 save_pretrained/tokenization failure are different speech-model issues." + "reason": "FastSpeech2ConformerTokenizer loading failure and Wav2Vec2 save_pretrained/tokenization failure are different model-specific problems." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug and Bamba torch.compile/SDPA failure are unrelated." + "reason": "SwitchTransformers sparse-layer construction and Bamba torch.compile SDPA failure are unrelated code paths." }, { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Qwen2.5-VL video position-id bug and FP8 inference failure on B200 are different problems." - }, - { - "left": "issue:43611", - "right": "issue:47405", - "accept": false, - "reason": "base_model_prefix loading regression and silent model_type mismatch conversion skip are separate model-loading bugs." - }, - { - "left": "issue:44898", - "right": "issue:45412", - "accept": false, - "reason": "Perceiver interpolation failure and RT-DETR memory leak are entirely different issues." + "reason": "Qwen2.5-VL video position_id bug is unrelated to FP8 inference failure on B200." }, { "left": "issue:45200", "right": "issue:46899", "accept": false, - "reason": "Gemma4 token-type defaults versus Gemma4 4-bit dtype cast corruption are both multimodal, but they affect different tensors and fixes." + "reason": "Gemma4 token-type-id defaults and 4-bit multimodal dtype casting are different quantization/input bugs." }, { - "left": "issue:44610", - "right": "issue:44857", + "left": "issue:43611", + "right": "issue:47405", "accept": false, - "reason": "Processor output-size mismatch and AMP float16 crash in LwDetrImageLoss are unrelated." + "reason": "base_model_prefix loading breakage and silent conversion skip on model_type mismatch are distinct loading regressions." } ] }, @@ -10106,6 +10182,7 @@ "issue:37428", "issue:38175", "issue:41720", + "issue:41950", "issue:42490", "issue:43335", "issue:43425", @@ -10119,253 +10196,259 @@ "issue:43854", "issue:43901", "issue:44112", - "issue:44206", - "issue:44617", + "issue:44292", + "issue:44610", "issue:44792", "issue:44843", "issue:44855", "issue:44857", + "issue:44898", "issue:45083", "issue:45200", "issue:45412", + "issue:45520", "issue:45563", - "issue:45584", "issue:45593", "issue:45759", "issue:45923", "issue:46209", "issue:46285", - "issue:46424", "issue:46431", "issue:46512", "issue:46566", - "issue:46697", "issue:46821", - "issue:46858", "issue:46932", "issue:47405", "issue:47427", - "issue:47436" + "issue:47436", + "issue:47448" ], "soft_pairs": [ + "issue:44610|issue:44857", "issue:37428|issue:43425", - "issue:45200|issue:46932", + "issue:44898|issue:45412", "issue:44843|issue:47427", + "issue:45200|issue:46932", "issue:43819|issue:45083", "issue:43335|issue:43653", - "issue:43606|issue:45083", "issue:41720|issue:46209", - "issue:44792|issue:45563", + "issue:43606|issue:45083", "issue:44857|issue:45593", + "issue:44792|issue:45563", + "issue:46431|issue:46566", "issue:43525|issue:47436", - "issue:43335|issue:43504", - "issue:28282|issue:33357", + "issue:41950|issue:47448", "issue:46512|issue:46821", - "issue:43550|issue:44112", + "issue:43335|issue:43504", "issue:45923|issue:47405", - "issue:38175|issue:45412", + "issue:28282|issue:33357", "issue:45759|issue:46285", - "issue:44206|issue:45584", - "issue:43827|issue:43901", + "issue:43550|issue:44112", "issue:29942|issue:32090", - "issue:46431|issue:46566", + "issue:43827|issue:43901", "issue:44855|issue:45083", "issue:43335|issue:43606", + "issue:38175|issue:45412", + "issue:44292|issue:45520", "issue:43335|issue:44112", - "issue:46424|issue:46858", - "issue:42490|issue:43854", - "issue:44617|issue:46697" + "issue:42490|issue:43854" ], - "prepared_review_unit_hash": "ef2e05f3d2d1d614a04d8ca5c940e2ef87264fdb18bcc4d47b223cd6ff7d3725", + "prepared_review_unit_hash": "c036f11a1fc5d117ac9fcc2eb4fda95fc44e8d17775160006c49888874e045b7", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35544, - "estimated_input_tokens": 8886, - "estimated_eval_tokens": 18028 + "soft_pair_count": 28, + "serialized_chars": 35745, + "estimated_input_tokens": 8937, + "estimated_eval_tokens": 18130 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the paired items share at most broad topical words, but not the same underlying bug or change. I would not collapse any of the proposed soft edges, and there is no strong single canonical artifact for the whole set.", - "confidence": 0.96, - "canonical_issue_reason": "No clear canonical issue: the items span unrelated failures across model loading, generation, docs, device/runtime backends, and model-specific regressions.", - "canonical_pr_reason": "No PRs are present in this cluster.", - "best_issue_reason": "None of the issues cleanly represents the whole cluster; if a representative must be chosen, it would still be misleading because the cluster is not a single defect family.", - "best_pr_reason": "No PRs are present in this cluster.", + "summary": "This cluster is mostly a set of unrelated issues that only look similar at the title/token level. None of the soft-edge pairs appears to be the same underlying bug or change, so all soft candidates should be rejected. I picked issue 43335 as the canonical issue only because it is the most connected node in the similarity graph, not because it truly represents a duplicate group.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43335 has the highest soft-edge connectivity in this cluster and is a concrete, well-scoped model bug, but it is only an arbitrary representative because the cluster is not actually a coherent duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43335 is the best available representative by graph centrality and specificity; however, the surrounding items are too heterogeneous for a strong canonical choice.", + "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different bugs in different code paths: OmDet image-size preprocessing mismatch vs LwDetr float16/CUDA loss crash." + }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "Both relate to Torch/import compatibility, but one is a missing flash-attn helper import and the other is a Torch 2.10 incompatibility; different failures and fixes." + "reason": "An import error from a missing internal symbol is not the same as a Torch 2.10 compatibility issue." }, { - "left": "issue:45200", - "right": "issue:46932", + "left": "issue:44898", + "right": "issue:45412", "accept": false, - "reason": "Both mention Gemma 4 dtype-related behavior, but one is about missing mm_token_type_ids defaults and the other is quantized vision/audio embedder casting; not the same bug." + "reason": "Perceiver non-default resolution handling and RT-DETR model memory release are unrelated model-specific problems." }, { "left": "issue:44843", "right": "issue:47427", "accept": false, - "reason": "Both concern initialization/model loading internals, but one is HF Hub offline behavior from an unconditional model_info() call and the other is redundant init skipping on non-CUDA hardware." + "reason": "Offline tokenizer patching via hub access is unrelated to redundant initialization checks on non-CUDA hardware." + }, + { + "left": "issue:45200", + "right": "issue:46932", + "accept": false, + "reason": "Both mention Gemma 4, but one is about missing text-only token type defaults and the other about quantized embedder dtype casting." }, { "left": "issue:43819", "right": "issue:45083", "accept": false, - "reason": "Different code paths and symptoms: DAC latent reconstruction mismatch vs a qwen3_omni_moe helper length calculation behavior." + "reason": "DAC latent reconstruction mismatch and qwen3_omni_moe feature-length helper behavior are different audio-model bugs." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "Both involve tokenizer/config registration issues, but one is sparse-layer construction in SwitchTransformersConfig and the other is BigBirdTokenizer special-token registration/decoding." + "reason": "Sparse-layer construction in SwitchTransformers is unrelated to BigBird tokenizer special-token registration." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Same broad Qwen family, but different failures: auto device mapping cuda assert vs FP8 inference on B200 hardware." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "Different subsystems: CPU offload device mismatch in a model vs a feature-extraction length helper behavior." + "reason": "CPU offload device mismatch in Bark is unrelated to qwen3_omni_moe output-length calculation." }, { - "left": "issue:41720", - "right": "issue:46209", + "left": "issue:44857", + "right": "issue:45593", "accept": false, - "reason": "Both are device/backend failures, but the concrete problems differ: auto device mapping on A800 vs FP8 inference on B200." + "reason": "Different model/components and different failure modes: AMP crash vs missing auxiliary losses." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "One is a janus image-generation test failure, the other is a stale generate() warning for paged generation; unrelated defects." + "reason": "janus image-generation test failure and a stale generation warning are not the same issue." }, { - "left": "issue:44857", - "right": "issue:45593", + "left": "issue:46431", + "right": "issue:46566", "accept": false, - "reason": "Different models and causes: LwDetr AMP/CUDA crash vs D-FINE skipping auxiliary losses when denoising is off." + "reason": "Chat-template availability in a processor and weights stuck on meta under device_map='auto' are unrelated." }, { "left": "issue:43525", "right": "issue:47436", "accept": false, - "reason": "Both are missing-config-attribute errors, but the missing attributes and affected model families differ (Llama4 pad_token_id vs Qwen2_5Omni max_position_embeddings)." - }, - { - "left": "issue:43335", - "right": "issue:43504", - "accept": false, - "reason": "Both are config/model preset bugs, but one is sparse-layer creation and the other is a legacy field preventing pretrained BeitForSemanticSegmentation load." + "reason": "Both are missing-config-attribute reports, but they affect different models and different attributes." }, { - "left": "issue:28282", - "right": "issue:33357", + "left": "issue:41950", + "right": "issue:47448", "accept": false, - "reason": "One is a PyTorch-not-installed ImportError for AutoModel; the other is a MacOS bus error on a specific CLIP community model." + "reason": "Video pipeline processor selection and Gemma4 architecture recognition are unrelated subsystems." }, { "left": "issue:46512", "right": "issue:46821", "accept": false, - "reason": "Both concern API/config behavior, but one is stale config.dtype after from_config(dtype=...) and the other is a window_size attribute error in TimesFM 2.5." + "reason": "A stale dtype field after from_config and a TimesFM window_size AttributeError are unrelated config bugs." }, { - "left": "issue:43550", - "right": "issue:44112", + "left": "issue:43335", + "right": "issue:43504", "accept": false, - "reason": "Both are CI/runtime regressions, but one is torch.compile with SDPA on Bamba-9B-v2 and the other is a stale device override test in GraniteSpeech." + "reason": "SwitchTransformers sparse-layer creation and Beit pretrained preset loading are different model-specific defects." }, { "left": "issue:45923", "right": "issue:47405", "accept": false, - "reason": "One is a missing gradient_checkpointing flag on a trust_remote_code variant; the other is silent conversion skip when loading a checkpoint through the wrong Auto class." + "reason": "Missing gradient-checkpointing metadata on a trust_remote_code model is unrelated to silent conversion skipping under the wrong Auto class." }, { - "left": "issue:38175", - "right": "issue:45412", + "left": "issue:28282", + "right": "issue:33357", "accept": false, - "reason": "Different problems: SigLIP2 zero probabilities vs RT-DETR memory not being released on deletion." + "reason": "A PyTorch-missing import error and a MacOS bus error in CLIP loading are different environment/runtime failures." }, { "left": "issue:45759", "right": "issue:46285", "accept": false, - "reason": "One is AutoModelForCausalLM.from_config failing to unwrap text_config for multimodal Qwen configs; the other is a structural inconsistency between CLIP text model classes." + "reason": "Both involve composite-model structure, but one is AutoModelForCausalLM text_config unwrapping and the other is CLIP class structural inconsistency." }, { - "left": "issue:44206", - "right": "issue:45584", + "left": "issue:43550", + "right": "issue:44112", "accept": false, - "reason": "Both are generation/processing regressions, but one is an unsupported center arg in LasrFeatureExtractor and the other is Whisper failing on empty transcription after align_special_tokens." + "reason": "Bamba torch.compile/SDPA failure and GraniteSpeech stale device override are unrelated model/runtime bugs." }, { - "left": "issue:43827", - "right": "issue:43901", + "left": "issue:29942", + "right": "issue:32090", "accept": false, - "reason": "Both are docs-related, but they target different APIs and mismatches: pipeline() removal docs vs return_all_scores docs." + "reason": "Flash Attention 2 test failures and NoneType GPU broadcast errors in Trainer do not share the same root cause." }, { - "left": "issue:29942", - "right": "issue:32090", - "accept": false, - "reason": "Different failures in different paths: Flash Attention 2 test failures vs a NoneType broadcast error in Trainer." - }, - { - "left": "issue:46431", - "right": "issue:46566", + "left": "issue:43827", + "right": "issue:43901", "accept": false, - "reason": "Different features: processor chat template validation vs DiffusionGemma weights staying on meta with device_map='auto'." + "reason": "Docs for pipeline removals and text-classification return_all_scores docs drift are separate documentation issues." }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "One is a Python 3.13 import/parsing issue with torch.jit.script decorators; the other is a qwen3_omni_moe helper length bug." + "reason": "A Python 3.13 torch.jit parsing failure is unrelated to qwen3_omni_moe feature-length behavior." }, { "left": "issue:43335", "right": "issue:43606", "accept": false, - "reason": "Sparse-layer config construction and CPU offload device mismatch are unrelated bugs." + "reason": "SwitchTransformers sparse-layer logic and Bark CPU offload device mismatch are unrelated bugs." }, { - "left": "issue:43335", - "right": "issue:44112", + "left": "issue:38175", + "right": "issue:45412", "accept": false, - "reason": "Sparse-layer config creation and a stale GraniteSpeech CI device override test are not the same underlying issue." + "reason": "Zero-probabilities behavior in a SigLIP2 model and RT-DETR memory retention are unrelated." }, { - "left": "issue:46424", - "right": "issue:46858", + "left": "issue:44292", + "right": "issue:45520", "accept": false, - "reason": "Both involve cache length behavior, but one is max_cache_length only affecting quantized cache while the other is get_seq_length() returning a device tensor that breaks static-cache arithmetic." + "reason": "Qwen-3-8B-NVFP4 runtime failure and flash_attn import key errors are different problems at different layers." }, { - "left": "issue:42490", - "right": "issue:43854", + "left": "issue:43335", + "right": "issue:44112", "accept": false, - "reason": "Both are loading-related, but one is an import warning for trailing-slash local paths and the other is a unit-test failure loading a specific GLM-4.7-Flash model." + "reason": "SwitchTransformers sparse-layer creation and GraniteSpeech stale device-override CI failure are unrelated." }, { - "left": "issue:44617", - "right": "issue:46697", + "left": "issue:42490", + "right": "issue:43854", "accept": false, - "reason": "Different model families and symptoms: CUDA OOM for Sam3Video vs image over-padding in Swin2SR." + "reason": "A local-path trailing-slash warning and a model-loading unit-test failure for GLM-4.7-Flash are not the same bug." } ] }, @@ -10403,6 +10486,7 @@ "issue:42175", "issue:42490", "issue:43278", + "issue:43329", "issue:43335", "issue:43388", "issue:43526", @@ -10415,42 +10499,43 @@ "issue:43825", "issue:44112", "issue:44206", - "issue:44292", "issue:44351", "issue:44448", "issue:44493", "issue:44514", "issue:44521", + "issue:44617", "issue:44743", "issue:44779", "issue:44857", "issue:45072", "issue:45083", "issue:45290", - "issue:45310", - "issue:45478", - "issue:45520", + "issue:45584", "issue:45588", - "issue:45593", "issue:45823", "issue:45859", "issue:46326", - "issue:47000" + "issue:46424", + "issue:46697", + "issue:46858" ], "soft_pairs": [ - "issue:43388|issue:43526", - "issue:43335|issue:45083", - "issue:44292|issue:45520", + "issue:44617|issue:46697", + "issue:46424|issue:46858", + "issue:44206|issue:45584", "issue:43819|issue:44493", + "issue:43388|issue:43526", "issue:45588|issue:45859", + "issue:43335|issue:45083", + "issue:10000|issue:15354", "issue:43606|issue:44857", - "issue:44206|issue:45823", - "issue:37428|issue:44351", - "issue:45290|issue:46326", "issue:39692|issue:44493", + "issue:45290|issue:46326", + "issue:37428|issue:44351", + "issue:44206|issue:45823", "issue:39692|issue:43526", "issue:44448|issue:44743", - "issue:10000|issue:15354", "issue:34689|issue:35141", "issue:44514|issue:44521", "issue:43720|issue:44112", @@ -10458,228 +10543,230 @@ "issue:43701|issue:45083", "issue:29127|issue:32090", "issue:44448|issue:44779", + "issue:25251|issue:32090", "issue:43550|issue:45083", "issue:32090|issue:33357", + "issue:37428|issue:42175", "issue:42490|issue:43575", "issue:43278|issue:43825", - "issue:37428|issue:42175", - "issue:45310|issue:45478", "issue:29942|issue:35141", - "issue:25251|issue:32090", - "issue:45593|issue:47000" + "issue:43278|issue:43329" ], - "prepared_review_unit_hash": "56c55d89c9800b2c81d7bb48f8cc02fc28b514a98467c5ff7630d7572691b0c9", + "prepared_review_unit_hash": "8149e8fbede4b4298dc508751171ce55e508f68efeb093bfea694fe0d98bd53b", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35893, - "estimated_input_tokens": 8974, - "estimated_eval_tokens": 18204 + "soft_pair_count": 30, + "serialized_chars": 36395, + "estimated_input_tokens": 9099, + "estimated_eval_tokens": 18454 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous and does not form one duplicate cluster. The only clear duplicate pair is issue:45310 and issue:45478, which report the same Qwen3.5 MoE `from_pretrained` failure. All other soft pairs look like separate bugs/regressions or different code paths.", - "confidence": 0.94, - "canonical_issue_reason": "Use issue:45310 as the canonical issue for the only true duplicate pair: it is the earlier report, has slightly richer discussion, and its title/body already capture the Qwen3.5 MoE `from_pretrained` error; issue:45478 is effectively the duplicate follow-up.", + "summary": "This cluster is thematically broad but not duplicate-heavy: it mixes unrelated Transformers bugs across chat templates, caching, dtype/device mismatches, model loading, preprocessing, and CI/test failures. None of the soft-similarity pairs look like the same underlying bug or change.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 44521 is the best representative anchor because it is open, well-described, has inbound references, and concerns a concrete multimodal chat-template bug with clear reproduction details.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:45310 is the best issue to keep as the representative report because it is the earliest and most developed description of the duplicated Qwen3.5 MoE loading failure.", + "best_issue_reason": "Issue 44521 is the strongest global issue candidate: it has the clearest user-facing failure, active discussion, and a specific code path (`apply_chat_template` for multimodal inputs).", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43388", - "right": "issue:43526", + "left": "issue:44617", + "right": "issue:46697", "accept": false, - "reason": "Both mention label handling, but one is about `gather_for_metrics` dropping tuple labels in the last batch and the other is `BeitImageProcessorFast.reduce_labels` returning only one label; different subsystems and failure modes." + "reason": "One is a video-model CUDA OOM, the other is an image-processor padding bug; different subsystems and failure modes." }, { - "left": "issue:43335", - "right": "issue:45083", + "left": "issue:46424", + "right": "issue:46858", "accept": false, - "reason": "A SwitchTransformers config construction bug is unrelated to a Qwen3 Omni helper function length calculation issue." + "reason": "Both involve cache behavior, but one is a config limit bug and the other is a device-tensor arithmetic issue in `_prefill`." }, { - "left": "issue:44292", - "right": "issue:45520", + "left": "issue:44206", + "right": "issue:45584", "accept": false, - "reason": "These are different problems: a Qwen-3-8B-NVFP4 runtime error versus a Python 3.13 `flash_attn` import-utils KeyError." + "reason": "Different audio-related problems: unsupported feature-extractor arg vs empty-transcription generation failure." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "DAC latent/forward mismatch and unexpected position-id keys are unrelated bugs with different code paths." + "reason": "Different code paths: DAC latent conversion mismatch vs position-id key warnings in model loading." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both mention labels, but one is distributed metric gathering and the other is BEiT label reduction; not the same bug." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Both touch flash-attention/MoE areas, but one is an `s_aux=None` crash in sink-less models and the other is a CPU-offload incompatibility in a rotary embedding; not the same underlying fix." + "reason": "Both touch model internals, but one is a flash-attention helper crash on `s_aux=None` and the other is CPU-offload incompatibility in rotary embeddings." }, { - "left": "issue:43606", - "right": "issue:44857", + "left": "issue:43335", + "right": "issue:45083", "accept": false, - "reason": "A CPU-offload device mismatch in Bark is not the same as a float16 AMP crash in LwDetrImageLoss." + "reason": "Unrelated bugs in different model components: SwitchTransformers config creation vs qwen3_omni_moe feature-length helper behavior." }, { - "left": "issue:44206", - "right": "issue:45823", + "left": "issue:10000", + "right": "issue:15354", "accept": false, - "reason": "Unsupported `center` arg handling in a feature extractor is unrelated to Gemma4 PLE forward device-mapping mismatch." + "reason": "Different models and failures: DistilBERT issue vs TorchScript GeneratorExp export incompatibility." }, { - "left": "issue:37428", - "right": "issue:44351", + "left": "issue:43606", + "right": "issue:44857", "accept": false, - "reason": "Both are import errors, but they reference different missing symbols (`_flash_supports_window_size` vs `HybridCache`) and different breakages." + "reason": "Both are device/dtype-related, but they affect different models and distinct mechanisms (CPU offload mismatch vs AMP/CUDA loss crash)." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "A documentation/example bug plus quantization failure is not the same as a model position-id warning regression." }, { "left": "issue:45290", "right": "issue:46326", "accept": false, - "reason": "Both involve chat templating, but one crashes on assistant messages with tool calls and no content, while the other concerns `continue_final_message` for mistral_common tokenizers." + "reason": "Both involve chat templating, but one crashes on assistant tool-call messages and the other breaks `continue_final_message` for mistral_common tokenizers." }, { - "left": "issue:39692", - "right": "issue:44493", + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and different modules; not the same underlying break." + }, + { + "left": "issue:44206", + "right": "issue:45823", "accept": false, - "reason": "A broken documentation example is not the same as a widespread unexpected key warning on position ids." + "reason": "Different models and issues: unsupported feature-extractor arg vs Gemma4 device-mapping mismatch during forward." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "These are unrelated: a SigLIP2 docs/quantization example issue versus a BEiT fast processor label-reduction bug." + "reason": "Both are vision/data-processing related, but one is a SigLIP2 docs example and the other is BEiT fast processor label handling." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Pegasus v4/v5 output drift and Qwen3.5 recurrent-state reset are different model behaviors and different code paths." - }, - { - "left": "issue:10000", - "right": "issue:15354", - "accept": false, - "reason": "The titles point to different models and unrelated issues: German DistilBertModel versus TorchScript generator-expression export problems in ViT." + "reason": "Both are v5 behavioral regressions, but one concerns Pegasus output changes while the other is recurrent-state reset logic in qwen3 modular code." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Model loading regression for Llama 3.2 90B Vision is not the same bug as output embeddings being reinitialized after resizing token embeddings." + "reason": "Different problems: Llama 3.2 Vision loading regression vs token-embedding resize reinitialization after `post_init`." }, { "left": "issue:44514", "right": "issue:44521", "accept": false, - "reason": "Both are `apply_chat_template` problems, but one is a batched-input crash with `padding=False` and the other is all-zero assistant masks for multimodal inputs; separate symptoms and likely separate fixes." + "reason": "Both are chat-template issues, but one crashes on batched padded input and the other returns all-zero assistant masks; distinct bugs." }, { "left": "issue:43720", "right": "issue:44112", "accept": false, - "reason": "Packed weights not being unpacked during accelerate loading is unrelated to a stale device-override CI test failure." + "reason": "A BitNet loading bug and a GraniteSpeech CI test flake are unrelated." }, { "left": "issue:44857", "right": "issue:45072", "accept": false, - "reason": "These are different dtype-related issues in different models: a float16 AMP crash versus bfloat16 inference mismatches." + "reason": "Both involve dtype/device mismatches, but in different models and execution settings; not the same code-path defect." }, { "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "A resume-from-checkpoint key mismatch is not the same as the Qwen3 Omni feature-length helper bug." + "reason": "Checkpoint key mismatch and qwen3_omni_moe feature-length behavior are unrelated." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "LayoutLMv3 error-message clarity and a Trainer `_gpu_broadcast_one` NoneType failure are unrelated." + "reason": "Different failure classes: clearer error messaging for layoutlmv3 vs NoneType broadcast error in Trainer." }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Both are v5 regressions, but one is Pegasus output differences and the other is Deepseek tokenizer correctness; different components and likely different fixes." + "reason": "Both are output regressions, but one is Pegasus generation drift and the other is tokenizer corruption in Deepseek." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to `_gpu_broadcast_one` NoneType failures." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "A torch.compile/SDPA failure in Bamba is unrelated to the Qwen3 Omni helper-function behavior." + "reason": "torch.compile/SDPA failure in Bamba is unrelated to qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Trainer broadcast of `NoneType` and a MacOS bus error in a community CLIP model are separate failures." + "reason": "Trainer broadcast NoneType error and MacOS bus error in community CLIP are distinct issues." + }, + { + "left": "issue:37428", + "right": "issue:42175", + "accept": false, + "reason": "An import-symbol regression is unrelated to missing TensorFlow in a pip extras install." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "A local-path import warning and an OOM when loading Qwen2-57B-A14B-Instruct with tensor parallelism are not the same issue." + "reason": "Local-path trailing-slash warning and tensor-parallel OOM are different problems." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Embedding dtype drift during evaluation is unrelated to a misleading `pipeline()` translation-task message." - }, - { - "left": "issue:37428", - "right": "issue:42175", - "accept": false, - "reason": "An import error for a flash-attention helper and TensorFlow not being included in the `.[torch]` backend are unrelated packaging/runtime issues." - }, - { - "left": "issue:45310", - "right": "issue:45478", - "accept": true, - "reason": "These are effectively the same report: identical title, same Qwen3.5 MoE `from_pretrained` error, and no distinguishing evidence beyond being duplicate submissions." + "reason": "BF16/FP32 dtype drift during eval is unrelated to a pipeline error-message regression." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Failing Flash Attention 2 tests and token-embedding resize reinitialization are different problems." - }, - { - "left": "issue:25251", - "right": "issue:32090", - "accept": false, - "reason": "Pipeline `top_k` output nesting and Trainer broadcast failures are unrelated." + "reason": "Flash Attention 2 test failures do not match an embedding resize reinitialization bug." }, { - "left": "issue:45593", - "right": "issue:47000", + "left": "issue:43278", + "right": "issue:43329", "accept": false, - "reason": "D-FINE auxiliary-loss behavior and RTDetrLoss infeasible cost matrix errors are separate model/loss bugs." + "reason": "A dtype change bug and an undefined-variable multimodal token-count bug are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative: the only accepted duplicate pair is issue:45310 / issue:45478, which share the same title and bug description. The remaining pairwise rejections are reasonably justified by differing model/component/failure-mode cues." - }, - "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 }, @@ -10713,28 +10800,27 @@ "issue:43257", "issue:43278", "issue:43299", - "issue:43329", "issue:43381", "issue:43526", "issue:43873", "issue:43927", - "issue:44186", "issue:44360", "issue:44448", "issue:44464", "issue:44530", "issue:44561", "issue:44589", - "issue:44779", "issue:44855", "issue:44977", "issue:45127", - "issue:45198", "issue:45229", "issue:45245", "issue:45290", + "issue:45310", "issue:45381", "issue:45459", + "issue:45478", + "issue:45593", "issue:45698", "issue:45920", "issue:46192", @@ -10742,249 +10828,246 @@ "issue:46399", "issue:46424", "issue:46655", + "issue:47000", "issue:47276", "issue:47436" ], "soft_pairs": [ + "issue:45310|issue:45478", + "issue:45593|issue:47000", "issue:45459|issue:46234", - "issue:43278|issue:43329", "issue:27957|issue:34689", "issue:26413|issue:31501", - "issue:45127|issue:45245", - "issue:44360|issue:44561", "issue:45290|issue:45381", - "issue:29870|issue:32090", + "issue:45127|issue:45245", "issue:40444|issue:43526", - "issue:44464|issue:46424", + "issue:29870|issue:32090", + "issue:44360|issue:44561", + "issue:44530|issue:47276", "issue:44448|issue:44977", + "issue:44464|issue:46424", "issue:32090|issue:33290", - "issue:44530|issue:47276", - "issue:29870|issue:33290", - "issue:44855|issue:46399", - "issue:43278|issue:43381", "issue:44589|issue:47436", + "issue:44855|issue:46399", + "issue:29870|issue:33290", "issue:42490|issue:43873", - "issue:45290|issue:46655", + "issue:43278|issue:43381", "issue:44977|issue:45229", + "issue:45290|issue:46655", + "issue:45920|issue:46192", "issue:43257|issue:43299", "issue:29925|issue:34689", "issue:43927|issue:45698", - "issue:45920|issue:46192", "issue:29942|issue:30064", - "issue:44561|issue:44779", - "issue:43873|issue:45229", "issue:33357|issue:39692", - "issue:44186|issue:45198" + "issue:43873|issue:45229" ], - "prepared_review_unit_hash": "1c11a2485efe71ba8df0e190596144d45bf1fcdaed42a80e8548250da5faac83", + "prepared_review_unit_hash": "24c001cc9d1e68a22c8883ec146a88211dd60858296f3fda52ee3864537eb6ab", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36475, - "estimated_input_tokens": 9119, - "estimated_eval_tokens": 18494 + "soft_pair_count": 28, + "serialized_chars": 36028, + "estimated_input_tokens": 9007, + "estimated_eval_tokens": 18270 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of unrelated issues; none of the soft pairs look like true duplicates. The broadest representative issue is the v5 tokenizer regression affecting many models (#45920), but it is not a strong canonical for the whole set.", - "confidence": 0.79, - "canonical_issue_reason": "#45920 is the broadest, highest-impact issue in the set and has the strongest cross-model signal, so it is the best representative issue even though the cluster is not a tight duplicate group.", + "summary": "Only one soft edge looks like a true duplicate: the two Qwen3.5 MoE from_pretrained regression reports. The rest are related only at a broad subsystem level (training, generation, tokenizer, multimodal, deepspeed, etc.) and should stay separate.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45310 is the better canonical: it was opened first and has the fuller report, while issue:45478 is essentially the same Qwen3.5 MoE from_pretrained failure with a near-identical title.", "canonical_pr_reason": null, - "best_issue_reason": "#45920 best represents the cluster because it describes a wide regression affecting multiple model families, unlike the other items which are narrower and mostly unrelated.", + "best_issue_reason": "issue:45310 is the strongest representative issue in the only true duplicate pair; it most cleanly captures the underlying Qwen3.5 MoE loading regression.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45459", - "right": "issue:46234", + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Near-identical titles and the same concrete failure: transformers>=5.4.0 Qwen3.5 MoE from_pretrained error. Same underlying bug." + }, + { + "left": "issue:45593", + "right": "issue:47000", "accept": false, - "reason": "Tokenizer/protobuf error handling vs a missing import_utils helper are unrelated bugs." + "reason": "Both concern detection losses, but one is auxiliary-loss handling in D-FINE and the other is an infeasible cost matrix in RTDetrLoss; different bugs and code paths." }, { - "left": "issue:43278", - "right": "issue:43329", + "left": "issue:45459", + "right": "issue:46234", "accept": false, - "reason": "Eval-time dtype drift and multimodal video token counting are different code paths." + "reason": "Both touch import-utils/tokenizer plumbing, but one is protobuf decode error masking and the other is a missing availability helper; unrelated issues." }, { "left": "issue:27957", "right": "issue:34689", "accept": false, - "reason": "Flash Attention support for XLM-Roberta is unrelated to Llama 3.2 Vision model loading." + "reason": "Both are model-loading/runtime problems, but for different models and different failures. Too broad to treat as duplicates." }, { "left": "issue:26413", "right": "issue:31501", "accept": false, - "reason": "Checkpoint resume epoch_iterator failure and FineWeb training startup are not the same bug." + "reason": "Both affect training startup, but one is checkpoint resume/epoch iterator handling and the other is a FineWeb SLM training launch failure; not the same bug." }, { - "left": "issue:45127", - "right": "issue:45245", + "left": "issue:45290", + "right": "issue:45381", "accept": false, - "reason": "LoRA merge/extended-vocab collapse is unrelated to category-count limits." + "reason": "One is chat template tokenization with tool-call assistant messages; the other is Qwen2.5-VL video vision position ids. Different functionality." }, { - "left": "issue:44360", - "right": "issue:44561", + "left": "issue:45127", + "right": "issue:45245", "accept": false, - "reason": "DSA indexer activation issue and removal of an import helper break different subsystems." + "reason": "LoRA merging with tied embeddings is unrelated to the category-count limit error." }, { - "left": "issue:45290", - "right": "issue:45381", + "left": "issue:40444", + "right": "issue:43526", "accept": false, - "reason": "Chat template tool-call crash and Qwen2.5-VL video position ids are distinct multimodal bugs." + "reason": "Both involve vision processors, but one is iterable multi-image VL finetuning and the other is BeitImageProcessorFast label reduction; different code paths." }, { "left": "issue:29870", "right": "issue:32090", "accept": false, - "reason": "Bart generate() decoder mask bug and Trainer _gpu_broadcast_one NoneType error are unrelated." + "reason": "Bart generate() decoder mask initialization and Trainer _gpu_broadcast_one NoneType errors are separate bugs." }, { - "left": "issue:40444", - "right": "issue:43526", + "left": "issue:44360", + "right": "issue:44561", "accept": false, - "reason": "IterableDataset multi-image finetuning and BeitImageProcessorFast label reduction are unrelated." + "reason": "A missing ReLU in the DSA indexer is unrelated to removal of is_torch_fx_available breaking trust_remote_code models." }, { - "left": "issue:44464", - "right": "issue:46424", + "left": "issue:44530", + "right": "issue:47276", "accept": false, - "reason": "Compiled forward chunked generation mismatch and quantized-cache length handling differ materially." + "reason": "PagedAttentionCache group-type failure and a misleading synced_gpus warning are not the same underlying issue." }, { "left": "issue:44448", "right": "issue:44977", "accept": false, - "reason": "Pegasus v4/v5 output drift and Qwen3.5 flash-attention generation failure are unrelated." + "reason": "Different model regressions: Pegasus v4/v5 output changes versus Qwen3.5 flash-attention generation problems." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation with compiled forward and cache max length handling are distinct generation/cache behaviors." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "NoneType broadcast failure and Adafactor/DeepSpeed OOM are different training issues." + "reason": "A NoneType broadcast failure in Trainer is not the same as an Adafactor+DeepSpeed OOM." }, { - "left": "issue:44530", - "right": "issue:47276", + "left": "issue:44589", + "right": "issue:47436", + "accept": false, + "reason": "Float8 storage lookup failure and missing max_position_embeddings on Qwen2_5OmniConfig are unrelated." + }, + { + "left": "issue:44855", + "right": "issue:46399", "accept": false, - "reason": "PagedAttentionCache group-type crash and synced_gpus warning text are unrelated." + "reason": "Both mention decorators/registration, but one is a Python 3.13 parsing/indentation issue and the other is a Qwen3.5 kernel registration crash." }, { "left": "issue:29870", "right": "issue:33290", "accept": false, - "reason": "Bart generation masking and DeepSpeed Adafactor OOM are not the same underlying bug." + "reason": "Bart generation mask bug and DeepSpeed Adafactor OOM are not duplicates." }, { - "left": "issue:44855", - "right": "issue:46399", + "left": "issue:42490", + "right": "issue:43873", "accept": false, - "reason": "Python 3.13 TorchScript parsing failure and missing decorator registration are different import/runtime problems." + "reason": "Local-path import warning with trailing slash and quantization offloading problems are different concerns." }, { "left": "issue:43278", "right": "issue:43381", "accept": false, - "reason": "Evaluation dtype change and gradient checkpointing in eval mode are unrelated behaviors." - }, - { - "left": "issue:44589", - "right": "issue:47436", - "accept": false, - "reason": "Float8 storage lookup error and missing max_position_embeddings on Qwen2_5OmniConfig are unrelated." + "reason": "One is an eval-time dtype change for embeddings; the other is a restriction on using gradient checkpointing in eval mode. Related but not the same bug." }, { - "left": "issue:42490", - "right": "issue:43873", + "left": "issue:44977", + "right": "issue:45229", "accept": false, - "reason": "Trailing-slash model path warning and quantization offloading behavior are different issues." + "reason": "Qwen3.5 flash-attention generation and Gemma4 multi-GPU OOM are unrelated failures." }, { "left": "issue:45290", "right": "issue:46655", "accept": false, - "reason": "Assistant-tool-call chat template crash and SmolVLM video padding bug are separate multimodal paths." + "reason": "Chat-template tokenization with tool calls and SmolVLM video padding are different multimodal bugs." }, { - "left": "issue:44977", - "right": "issue:45229", + "left": "issue:45920", + "right": "issue:46192", "accept": false, - "reason": "Qwen3.5 flash-attention generation and Gemma4 multi-GPU OOM are unrelated." + "reason": "Wrong token IDs in AutoTokenizer and silent return_tensors ignoring are separate tokenizer API issues." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Accelerate/DeepSpeed MoE conversion failure and v5 Qwen3VL MoE loading regression are similar in theme but not the same concrete bug." + "reason": "Both involve Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is a broader from_pretrained failure for Qwen3VL MoE; not clearly the same bug." }, { "left": "issue:29925", "right": "issue:34689", "accept": false, - "reason": "SigLIP fast tokenizer enforcement and Llama 3.2 Vision loading are unrelated." + "reason": "SigLIP tokenizer fast-mode enforcement and Llama 3.2 Vision model loading break are unrelated." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Custom token IDs being lost and loading the wrong custom module after save are different serialization bugs." - }, - { - "left": "issue:45920", - "right": "issue:46192", - "accept": false, - "reason": "Wrong token IDs regression and return_tensors being ignored are separate tokenizer API issues." + "reason": "Config token-ID persistence and wrong custom-module loading after save_pretrained are both serialization-related, but they are different failure modes." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention test failures and void segmentation map processing are unrelated." + "reason": "Failing Flash Attention 2 tests and void segmentation-map processing are separate problems." }, { - "left": "issue:44561", - "right": "issue:44779", + "left": "issue:33357", + "right": "issue:39692", "accept": false, - "reason": "Removed helper breakage for trust_remote_code models and DeepSeek tokenizer regression are not the same issue." + "reason": "MacOS bus error on community CLIP loading and SigLIP2 documentation errors are different issues." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Quantization offloading behavior and Gemma4 multi-GPU memory exhaustion are different problems." - }, - { - "left": "issue:33357", - "right": "issue:39692", - "accept": false, - "reason": "MacOS bus error on CLIP loading and SigLIP2 documentation example errors are not the same bug." - }, - { - "left": "issue:44186", - "right": "issue:45198", - "accept": false, - "reason": "LayoutLMv2 tokenization/padding crash and Wav2Vec2 save/tokenization failure are unrelated." + "reason": "Quantization/offloading behavior and Gemma4 multi-GPU OOM are not the same code-path problem." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The conclusions are grounded in the packet and the duplicate call is conservative: issue:45310 and issue:45478 have near-identical titles and clearly point to the same Qwen3.5 MoE from_pretrained regression. The remaining pairs are rejected with appropriately cautious, title-level distinctions rather than overstated duplicate claims." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -11007,7 +11090,6 @@ "issue:34567", "issue:34689", "issue:41720", - "issue:42490", "issue:42915", "issue:42994", "issue:43065", @@ -11024,6 +11106,7 @@ "issue:43976", "issue:43994", "issue:44077", + "issue:44186", "issue:44206", "issue:44265", "issue:44361", @@ -11032,15 +11115,14 @@ "issue:44448", "issue:44466", "issue:44479", - "issue:44568", + "issue:44561", "issue:44743", + "issue:44779", "issue:44871", "issue:44977", "issue:45198", "issue:45229", - "issue:45412", "issue:45440", - "issue:45491", "issue:45563", "issue:45588", "issue:45803", @@ -11052,251 +11134,230 @@ ], "soft_pairs": [ "issue:45588|issue:45823", + "issue:44186|issue:45198", + "issue:44561|issue:44779", "issue:44448|issue:44466", - "issue:23647|issue:29870", + "issue:43412|issue:45588", "issue:44871|issue:44977", + "issue:23647|issue:29870", "issue:43381|issue:43701", "issue:46421|issue:46858", "issue:44361|issue:45198", "issue:29127|issue:34567", - "issue:44206|issue:44479", - "issue:43278|issue:43575", "issue:41720|issue:45229", - "issue:43412|issue:45588", + "issue:43278|issue:43575", + "issue:43334|issue:43531", "issue:44265|issue:45198", - "issue:43976|issue:45440", - "issue:43611|issue:46121", + "issue:44206|issue:44479", "issue:44977|issue:45563", "issue:43653|issue:44442", + "issue:43611|issue:46121", + "issue:42915|issue:42994", + "issue:43976|issue:45440", "issue:45198|issue:45803", + "issue:43065|issue:43994", "issue:29870|issue:34689", - "issue:42915|issue:42994", "issue:44077|issue:47276", - "issue:43334|issue:43531", "issue:43653|issue:44743", - "issue:43065|issue:43994", "issue:43756|issue:45440", - "issue:44568|issue:45491", - "issue:43575|issue:43701", - "issue:43575|issue:44387", - "issue:43994|issue:45412", - "issue:43065|issue:45412", - "issue:42490|issue:43381" + "issue:43575|issue:44387" ], - "prepared_review_unit_hash": "4d06f086d57302c2a7e7b12c914c6c97d7bcf893360411696144e4b7f2ba2c6b", + "prepared_review_unit_hash": "cd2e5c61b45e07f9831539742959e03d8b1a2f93cf2c3ea6b36292ab159f3a07", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37133, - "estimated_input_tokens": 9284, - "estimated_eval_tokens": 18824 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35569, + "estimated_input_tokens": 8893, + "estimated_eval_tokens": 18042 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Loose, heterogeneous cluster of unrelated issue reports across many subsystems (tokenizers, generation, cache/compile, quantization, vision, training state). None of the soft-similarity pairs look like the same underlying bug, so no duplicates should be merged.", - "confidence": 0.95, - "canonical_issue_reason": "No canonical issue: the items are not one duplicate set, but a grab-bag of unrelated bugs and regressions.", + "summary": "This cluster is heterogeneous: it mixes unrelated tokenizer, generation, quantization, cache, export, and training/runtime issues across different models and code paths. None of the soft pairs look like the same underlying bug/change, so no duplicate merge is warranted.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out because the cluster is not a coherent duplicate set; the items span distinct subsystems and failure modes.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a good representative for the cluster because the reports do not share one concrete failure mode.", + "best_issue_reason": "No global best issue candidate is appropriate here because these reports are not variants of one problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45588", "right": "issue:45823", "accept": false, - "reason": "Different bugs: flash_attention crashes on s_aux=None, while the other is a Gemma4 device-mapping mismatch." + "reason": "Both are forward/inference bugs, but one is a flash-attention AttributeError and the other is a device-map mismatch in Gemma4; different code paths and fixes." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizers, different models, and different failure modes (NER/padding crash vs save_pretrained/tokenization failure)." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "One is a v5 API removal breaking trust_remote_code models; the other is a DeepSeek tokenizer regression. Not the same bug." }, { "left": "issue:44448", "right": "issue:44466", "accept": false, - "reason": "One reports output differences between v4/v5; the other is inconsistent lm_head.weight serialization by device." + "reason": "Both are v5 regressions, but one changes Pegasus generation output while the other is about lm_head.weight serialization; distinct underlying issues." }, { - "left": "issue:23647", - "right": "issue:29870", + "left": "issue:43412", + "right": "issue:45588", "accept": false, - "reason": "W&B sweeps custom objective and BART generate decoder_attention_mask crash are unrelated." + "reason": "Export-time self-training-guard/erfinv problems versus a runtime flash-attention AttributeError; not mergeable as one fix." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma eos_token_id config mismatch and Qwen3.5 flash-attention generation failure are distinct issues." + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:23647", + "right": "issue:29870", + "accept": false, + "reason": "WandB sweep custom objective integration and BART generate decoder_attention_mask initialization are unrelated." }, { "left": "issue:43381", "right": "issue:43701", "accept": false, - "reason": "Gradient checkpointing not allowed in eval mode is unrelated to resume_from_checkpoint key mismatch." + "reason": "Eval-mode gradient checkpointing restriction and resume_from_checkpoint key mismatch are separate trainer issues." }, { "left": "issue:46421", "right": "issue:46858", "accept": false, - "reason": "Both involve cache/prefill, but one is lazy-init recompilation and the other is a device-tensor arithmetic bug." + "reason": "Both touch cache/compile behavior, but one is lazy-init recompilation and the other is a device-tensor seq_length arithmetic bug; different concrete defects." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "Different tokenizers/models and different failure paths; not the same underlying bug." + "reason": "Different tokenizer/model families and different symptoms; no shared code-path bug is evident." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error clarity and TrainerState token counting are unrelated." + "reason": "A LayoutLMv3 error-message clarity issue is unrelated to TrainerState token counting not updating." }, { - "left": "issue:44206", - "right": "issue:44479", + "left": "issue:41720", + "right": "issue:45229", "accept": false, - "reason": "Unsupported center arg in an audio feature extractor is not the same as the Qwen VL video regression." + "reason": "Both involve large-model inference resource problems, but cudaErrorAssert on Qwen3 is not the same bug as Gemma4 multi-GPU OOM." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "Eval dtype drift and tensor-parallel OOM are different problems." - }, - { - "left": "issue:41720", - "right": "issue:45229", - "accept": false, - "reason": "Both are multi-GPU/memory related, but the models and concrete failures differ (assert vs OOM)." + "reason": "Embedding dtype drift during evaluate and Qwen2 tensor-parallel OOM are unrelated." }, { - "left": "issue:43412", - "right": "issue:45588", + "left": "issue:43334", + "right": "issue:43531", "accept": false, - "reason": "ExecuTorch export/self-training-guard failure is unrelated to the flash_attention s_aux=None crash." + "reason": "A pad_token_id attribute error in Qwen3-VL is not the same issue as sliding_window handling in Qwen3-MoE." }, { "left": "issue:44265", "right": "issue:45198", "accept": false, - "reason": "torch.export with torch_compilable_check and Wav2Vec2 save/tokenization failure are separate bugs." - }, - { - "left": "issue:43976", - "right": "issue:45440", - "accept": false, - "reason": "Python version compatibility regression and DeepSeekV3 implementation divergence are unrelated." + "reason": "torch.export failures with torch_compilable_check are unrelated to Wav2Vec2 save/tokenization failures." }, { - "left": "issue:43611", - "right": "issue:46121", + "left": "issue:44206", + "right": "issue:44479", "accept": false, - "reason": "base_model_prefix loading regression and rope param dict TypeError are different code paths." + "reason": "Both are regressions, but one is a feature extractor arg mismatch and the other is a video-input regression in Qwen VL models; different fixes." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and stale paged-generate warning are not the same issue." + "reason": "Flash-attention generation failure and a stale paged-generate warning are different problems in generation code." }, { "left": "issue:43653", "right": "issue:44442", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug and FastSpeech2ConformerTokenizer load failure are distinct." - }, - { - "left": "issue:45198", - "right": "issue:45803", - "accept": false, - "reason": "Wav2Vec2 save/tokenization failure is unrelated to the bare-except problem in FuyuBatchFeature." + "reason": "BigBirdTokenizer special-token registration and FastSpeech2Conformer tokenizer loading are distinct tokenizer issues." }, { - "left": "issue:29870", - "right": "issue:34689", + "left": "issue:43611", + "right": "issue:46121", "accept": false, - "reason": "BART decoder mask generate crash and Llama 3.2 Vision model-loading regression are unrelated." + "reason": "Model loading with base_model_prefix and rope-parameter dict conversion are unrelated configuration/validation bugs." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "Both touch quantization, but one is FineGrainedFP8Config failure and the other is quantized-model saving failure." - }, - { - "left": "issue:44077", - "right": "issue:47276", - "accept": false, - "reason": "PatchTSMixer post_init policy and synced_gpus warning wording are unrelated." + "reason": "FineGrainedFP8Config failure and quantized model saving failure are different quantization-related workflows, not the same defect." }, { - "left": "issue:43334", - "right": "issue:43531", + "left": "issue:43976", + "right": "issue:45440", "accept": false, - "reason": "Qwen3-VL pad_token_id load error and Qwen3-MoE sliding_window issue are different bugs." + "reason": "A Python version compatibility regression and a DeepSeekV3MoE implementation divergence are unrelated." }, { - "left": "issue:43653", - "right": "issue:44743", + "left": "issue:45198", + "right": "issue:45803", "accept": false, - "reason": "Tokenizer special-token decode bug and recurrent-state reset in Qwen3.5 cache are unrelated." + "reason": "Wav2Vec2 tokenization/save failures and a bare-except in FuyuBatchFeature are separate issues." }, { "left": "issue:43065", "right": "issue:43994", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder and SigLIP2 nonsensical outputs are different model issues." - }, - { - "left": "issue:43756", - "right": "issue:45440", - "accept": false, - "reason": "Different models and failure modes: Smollm3 RoPE-layer mismatch vs DeepSeekV3MoE divergence." - }, - { - "left": "issue:44568", - "right": "issue:45491", - "accept": false, - "reason": "Tokenizer special-token insertion bug and Gemma3 NaN embeddings with sliding window are not the same bug." + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to SigLIP2 producing nonsensical outputs through AutoModel/pipeline." }, { - "left": "issue:43575", - "right": "issue:43701", + "left": "issue:29870", + "right": "issue:34689", "accept": false, - "reason": "Tensor-parallel load OOM is unrelated to resume_from_checkpoint key mismatch." + "reason": "BART generate UnboundLocalError and Llama 3.2 Vision model loading breakage are different code paths." }, { - "left": "issue:43575", - "right": "issue:44387", + "left": "issue:44077", + "right": "issue:47276", "accept": false, - "reason": "Both mention memory, but one is TP load OOM and the other is int4 reserved-memory regression; not the same code-path problem." + "reason": "Optional post_init validation for patchtsmixer is unrelated to the synced_gpus warning text in continuous batching." }, { - "left": "issue:43994", - "right": "issue:45412", + "left": "issue:43653", + "right": "issue:44743", "accept": false, - "reason": "SigLIP2 wrong outputs and RT-DETR memory not released are unrelated." + "reason": "Tokenizer special-token registration and Qwen3.5 cache state reset behavior are unrelated." }, { - "left": "issue:43065", - "right": "issue:45412", + "left": "issue:43756", + "right": "issue:45440", "accept": false, - "reason": "Dummy layer in Sam3PixelDecoder is unrelated to an RT-DETR memory leak." + "reason": "Smollm3 RoPE-layer dropping and DeepSeekV3MoE divergence are different model implementation bugs." }, { - "left": "issue:42490", - "right": "issue:43381", + "left": "issue:43575", + "right": "issue:44387", "accept": false, - "reason": "Trailing-slash import warning and gradient-checkpointing eval-mode restriction are unrelated." + "reason": "Tensor-parallel OOM for Qwen2 is not the same as int4 quantization increasing CUDA reserved memory in Transformers 5.x." } ] }, @@ -11324,17 +11385,18 @@ "issue:27957", "issue:28282", "issue:29127", - "issue:29870", - "issue:29925", "issue:33290", "issue:33357", "issue:34689", "issue:35141", + "issue:42490", "issue:42915", - "issue:43066", + "issue:43065", "issue:43381", "issue:43421", "issue:43526", + "issue:43575", + "issue:43701", "issue:43824", "issue:43873", "issue:43931", @@ -11342,10 +11404,10 @@ "issue:44186", "issue:44279", "issue:44442", - "issue:44462", + "issue:44568", + "issue:44610", "issue:44625", "issue:44857", - "issue:44898", "issue:45072", "issue:45245", "issue:45250", @@ -11365,263 +11427,262 @@ "issue:46640", "issue:46693", "issue:46701", - "issue:46821", "issue:47217" ], "soft_pairs": [ + "issue:44568|issue:45491", + "issue:43575|issue:43701", + "issue:42490|issue:43381", "issue:29127|issue:33357", + "issue:43994|issue:45412", "issue:45491|issue:46693", "issue:45372|issue:45588", "issue:44442|issue:45072", "issue:45406|issue:46701", - "issue:43526|issue:47217", + "issue:43065|issue:45412", "issue:44625|issue:45584", - "issue:45440|issue:45593", + "issue:45250|issue:46640", + "issue:43526|issue:47217", + "issue:43526|issue:45834", "issue:29127|issue:35141", + "issue:45440|issue:45593", "issue:45459|issue:45712", - "issue:43421|issue:45491", - "issue:43526|issue:45834", - "issue:45250|issue:46640", - "issue:45412|issue:46369", "issue:45372|issue:45406", - "issue:45245|issue:45563", - "issue:45372|issue:45459", - "issue:43824|issue:44279", "issue:43994|issue:46369", "issue:43526|issue:44186", + "issue:43824|issue:44279", "issue:43381|issue:43873", + "issue:43421|issue:45491", + "issue:45372|issue:45459", + "issue:42915|issue:43931", "issue:33290|issue:35141", "issue:33290|issue:34689", + "issue:45245|issue:45563", "issue:27957|issue:33290", - "issue:42915|issue:43931", - "issue:43066|issue:44462", - "issue:23647|issue:28282", "issue:44857|issue:45588", - "issue:45588|issue:46821", - "issue:29870|issue:29925", - "issue:27957|issue:29127", - "issue:44442|issue:44898" + "issue:23647|issue:28282", + "issue:44610|issue:45412" ], - "prepared_review_unit_hash": "6477fa142a0402259441aeb89d458ee67fa6c53da0d56b055405c81567ff2360", + "prepared_review_unit_hash": "a9f0039276d15d1dec458a78b3788e14cbc0dcf7f14910ca518ec7b13286c152", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 36507, - "estimated_input_tokens": 9127, - "estimated_eval_tokens": 18510 + "serialized_chars": 36484, + "estimated_input_tokens": 9121, + "estimated_eval_tokens": 18498 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the items share broad Transformers-adjacent themes (tokenizers, model loading, attention, generation), but none of the soft pairs look like the same underlying bug/change. I would not merge any of the proposed issue pairs.", - "confidence": 0.11, - "canonical_issue_reason": "No true canonical issue emerges because the cluster is not a duplicate set; if a representative is needed, issue 45372 is the clearest concrete regression report, but it does not subsume the others.", + "summary": "No soft pair is a safe duplicate: these are mostly unrelated bug reports, regressions, and feature requests across different models and code paths. Best standalone issue is the FA2 long-context regression because it is specific, actionable, and clearly scoped.", + "confidence": 0.94, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45372 is the strongest representative only in the sense that it has a specific, reproducible failure mode and clear user impact; it is not a duplicate of the rest of the cluster.", + "best_issue_reason": "issue:46693 is the strongest standalone issue: it describes a concrete Flash Attention 2 performance regression with long-context generation, a clear repro surface, and actionable scope.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Different failures and models: tokenizer BOS/EOS behavior vs Gemma3 NaN embeddings in mixed-length batching." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "One is TP memory/OOM on Qwen2-57B; the other is a checkpoint key mismatch during resume. Not the same bug." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Import warning for local model paths is unrelated to gradient checkpointing being disallowed in eval mode." + }, { "left": "issue:29127", "right": "issue:33357", "accept": false, - "reason": "Different subsystems and failures: LayoutLMv3 error-message clarity vs CLIP bus error on macOS." + "reason": "LayoutLMv3 box validation messaging and a CLIP bus error on macOS are different issues and code paths." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 bad outputs and RT-DETR memory not being released are unrelated model/runtime problems." }, { "left": "issue:45491", "right": "issue:46693", "accept": false, - "reason": "Both involve FA2/long-context language models, but one is NaN embeddings and the other is a performance regression." + "reason": "Both mention FA2-adjacent context, but one is NaN embeddings and the other is a performance regression; different underlying bugs." }, { "left": "issue:45372", "right": "issue:45588", "accept": false, - "reason": "Distinct bugs: Gemma 4 processor import/load failure vs flash_attention.py crashing on sink-less models." + "reason": "Gemma 4 import failure from mistral_common is not the same as a flash_attention s_aux=None AttributeError." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "Different code paths and symptoms: AutoTokenizer loading failure vs bfloat16 dtype mismatch in inference." + "reason": "AutoTokenizer loading failure and bfloat16 dtype mismatches are distinct problems." }, { "left": "issue:45406", "right": "issue:46701", "accept": false, - "reason": "Gemma4Processor serve crash is unrelated to the DiffusionGemmaGenerationOutput pipeline failure." + "reason": "Gemma4Processor serve crash from missing _tokenizer is separate from DiffusionGemma pipeline behavior." }, { - "left": "issue:43526", - "right": "issue:47217", + "left": "issue:43065", + "right": "issue:45412", "accept": false, - "reason": "BEiT fast image-processor label handling is unrelated to VL image-placeholder failures in user text." + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory leaks on deletion." }, { "left": "issue:44625", "right": "issue:45584", "accept": false, - "reason": "Qwen3.5 num_labels propagation and Whisper empty-transcription generation are separate bugs." - }, - { - "left": "issue:45440", - "right": "issue:45593", - "accept": false, - "reason": "DeepseekV3MoE divergence from remote impl is not the same as D-FINE losing auxiliary losses." + "reason": "num_labels propagation in Qwen3.5 is unrelated to Whisper failing on empty transcription after special-token alignment." }, { - "left": "issue:29127", - "right": "issue:35141", + "left": "issue:45250", + "right": "issue:46640", "accept": false, - "reason": "One is LayoutLMv3 error messaging; the other is embedding reinitialization after resizing token embeddings." + "reason": "Both are generic Flash Attention 2 topics, but neither identifies the same concrete bug or fixable code path." }, { - "left": "issue:45459", - "right": "issue:45712", + "left": "issue:43526", + "right": "issue:47217", "accept": false, - "reason": "Tokenizer error masking without protobuf is unrelated to leftover dummy classes leaking into dir(transformers)." + "reason": "BeitImageProcessorFast reduce_labels bug and VL image-placeholder failures are different subsystems and symptoms." }, { - "left": "issue:43421", - "right": "issue:45491", + "left": "issue:43526", + "right": "issue:45834", "accept": false, - "reason": "Runtime special-token post-processor updates and Gemma3 NaN embeddings are different defects." + "reason": "Reduce_labels returning one label and Kosmos2.5 OCR index errors are unrelated." }, { - "left": "issue:43526", - "right": "issue:45834", + "left": "issue:29127", + "right": "issue:35141", "accept": false, - "reason": "BEiT label reduction bug does not match Kosmos2.5 long-OCR index errors." + "reason": "LayoutLMv3 error messaging and token-embedding reinitialization on resize are not the same defect." }, { - "left": "issue:45250", - "right": "issue:46640", + "left": "issue:45440", + "right": "issue:45593", "accept": false, - "reason": "Both mention Flash Attention 2, but the issues are too vague and do not clearly describe the same bug." + "reason": "DeepseekV3MoE divergence from remote implementation is unrelated to D-FINE auxiliary loss handling." }, { - "left": "issue:45412", - "right": "issue:46369", + "left": "issue:45459", + "right": "issue:45712", "accept": false, - "reason": "RT-DETR memory leak on deletion is different from SAM3 LiteText memory overflow during call." + "reason": "Hidden tokenizer errors without protobuf and leftover dummy classes are separate maintenance/behavior issues." }, { "left": "issue:45372", "right": "issue:45406", "accept": false, - "reason": "Both involve Gemma 4, but one is an import dependency failure and the other is a missing _tokenizer attribute at serve time." + "reason": "One is an import-time dependency failure; the other is a runtime serve crash with a missing attribute." }, { - "left": "issue:45245", - "right": "issue:45563", + "left": "issue:43994", + "right": "issue:46369", "accept": false, - "reason": "Category-count RuntimeError and stale generate() warning are unrelated." + "reason": "Different models and failures: nonsensical SigLIP2 outputs vs Sam3-litetext memory overflow." }, { - "left": "issue:45372", - "right": "issue:45459", + "left": "issue:43526", + "right": "issue:44186", "accept": false, - "reason": "Gemma processor import failure and protobuf-related tokenizer error masking are separate problems." + "reason": "BeitImageProcessorFast reduce_labels and LayoutLMv2Tokenizer NER/padding crashes are unrelated." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "Missing Qwen2_5_VL import is not the same as a generic transformers dependency issue." + "reason": "Specific Qwen2_5_VL import error vs a vague dependency issue; not the same underlying bug." }, { - "left": "issue:43994", - "right": "issue:46369", + "left": "issue:43381", + "right": "issue:43873", "accept": false, - "reason": "SigLIP2 nonsensical outputs and SAM3 LiteText memory overflow are unrelated failures." + "reason": "Gradient checkpointing in eval mode and quantization offloading behavior are different code paths." }, { - "left": "issue:43526", - "right": "issue:44186", + "left": "issue:43421", + "right": "issue:45491", "accept": false, - "reason": "Different tokenizer/image-processor bugs affecting different tasks and failure modes." + "reason": "Runtime post-processor updates for special tokens do not match Gemma3 NaN embeddings." }, { - "left": "issue:43381", - "right": "issue:43873", + "left": "issue:45372", + "right": "issue:45459", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to quantization offloading behavior." + "reason": "Gemma4 processor import failure and protobuf-related tokenizer error masking are different problems." }, { - "left": "issue:33290", - "right": "issue:35141", + "left": "issue:42915", + "right": "issue:43931", "accept": false, - "reason": "Adafactor+DeepSpeed OOM is unrelated to post_init reinitializing resized embeddings." + "reason": "Qwen3Moe FP8 config failure and Qwen3-VL-30B weight-shape mismatch are distinct model-loading issues." }, { "left": "issue:33290", - "right": "issue:34689", - "accept": false, - "reason": "Different model/runtime failures: optimizer OOM versus Llama 3.2 Vision model loading breakage." - }, - { - "left": "issue:27957", - "right": "issue:33290", + "right": "issue:35141", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Adafactor OOM in DeepSpeed are unrelated issues." + "reason": "DeepSpeed Adafactor OOM is unrelated to embedding reinitialization after resizing." }, { - "left": "issue:42915", - "right": "issue:43931", + "left": "issue:33290", + "right": "issue:34689", "accept": false, - "reason": "Qwen3Moe FP8 config failure and Qwen3-VL weight-shape mismatch are different model-specific bugs." + "reason": "Optimizer OOM and Llama 3.2 Vision model loading breakage are different defects." }, { - "left": "issue:43066", - "right": "issue:44462", + "left": "issue:45245", + "right": "issue:45563", "accept": false, - "reason": "Wrong tokenizer decoder type in v5 is not the same as AutoTokenizer ignoring tokenizer.json." + "reason": "Category-cardinality runtime error and a stale generate warning are unrelated." }, { - "left": "issue:23647", - "right": "issue:28282", + "left": "issue:27957", + "right": "issue:33290", "accept": false, - "reason": "Wandb sweeps custom objective integration and missing PyTorch ImportError are unrelated." + "reason": "Flash Attention 2 support for XLMRoberta is not the same bug as Adafactor+DeepSpeed OOM." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "LwDetrImageLoss AMP/CUDA crash is a different failure than flash_attention.py's s_aux=None AttributeError." - }, - { - "left": "issue:45588", - "right": "issue:46821", - "accept": false, - "reason": "Flash-attention sink-less-model crash and TimesFM window_size AttributeError are unrelated." - }, - { - "left": "issue:29870", - "right": "issue:29925", - "accept": false, - "reason": "BART generate() decoder_attention_mask bug and SigLIP tokenizer fast-mode enforcement are unrelated." + "reason": "LwDetrImageLoss AMP CUDA crash and flash_attention s_aux=None AttributeError are separate failures." }, { - "left": "issue:27957", - "right": "issue:29127", + "left": "issue:23647", + "right": "issue:28282", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support is unrelated to LayoutLMv3 error-message clarity." + "reason": "Wandb sweeps custom objective support and missing PyTorch import errors do not share an underlying code-path." }, { - "left": "issue:44442", - "right": "issue:44898", + "left": "issue:44610", + "right": "issue:45412", "accept": false, - "reason": "AutoTokenizer loading FastSpeech2ConformerTokenizer and Perceiver non-default resolution failures are different bugs." + "reason": "OmDet-Turbo processor image-size mismatch and RT-DETR memory retention are unrelated issues." } ] }, @@ -11648,19 +11709,19 @@ "issue:25251", "issue:27957", "issue:28282", + "issue:29127", "issue:29870", + "issue:29925", "issue:42175", "issue:42490", "issue:42617", + "issue:43066", "issue:43262", "issue:43316", "issue:43493", - "issue:43525", - "issue:43550", "issue:43575", "issue:43643", "issue:43646", - "issue:43701", "issue:43720", "issue:43749", "issue:43761", @@ -11670,22 +11731,21 @@ "issue:44077", "issue:44206", "issue:44442", + "issue:44462", "issue:44479", "issue:44514", "issue:44743", "issue:44811", "issue:44871", + "issue:44898", "issue:44913", - "issue:44964", "issue:44977", - "issue:44987", "issue:45245", "issue:45381", - "issue:45440", + "issue:45588", "issue:45759", "issue:45859", "issue:46121", - "issue:46178", "issue:46234", "issue:46399", "issue:46821", @@ -11695,217 +11755,217 @@ "issue:47448" ], "soft_pairs": [ + "issue:43066|issue:44462", + "issue:45588|issue:46821", + "issue:29870|issue:29925", + "issue:25251|issue:28282", + "issue:44442|issue:44898", + "issue:27957|issue:29127", "issue:44514|issue:45381", "issue:46121|issue:46399", - "issue:43262|issue:44206", "issue:43761|issue:45859", - "issue:43316|issue:47312", "issue:43646|issue:44077", "issue:46821|issue:47436", + "issue:43316|issue:47312", "issue:43720|issue:44442", - "issue:25251|issue:28282", "issue:44743|issue:44811", - "issue:44977|issue:45245", "issue:42175|issue:46234", + "issue:43262|issue:44206", "issue:44077|issue:44479", "issue:43873|issue:44871", "issue:43316|issue:45759", "issue:42490|issue:44743", - "issue:43493|issue:43827", + "issue:44977|issue:45245", "issue:47276|issue:47448", + "issue:43493|issue:43827", "issue:43749|issue:43931", - "issue:43643|issue:44913", "issue:42617|issue:43575", - "issue:44987|issue:46178", - "issue:43316|issue:43525", - "issue:43550|issue:43701", - "issue:44964|issue:45440", - "issue:27957|issue:29870" + "issue:43643|issue:44913" ], - "prepared_review_unit_hash": "dfd46f2806bbb399919900bbedb164f4f9f6d7389bb6ca7995e7b75f45dc68ea", + "prepared_review_unit_hash": "ae71bdb3138621f6c338e99afd76e75f3e5b29de5a16850d1c6e3f776d473b88", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 26, - "serialized_chars": 35437, - "estimated_input_tokens": 8860, - "estimated_eval_tokens": 17976 + "serialized_chars": 34865, + "estimated_input_tokens": 8717, + "estimated_eval_tokens": 17690 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is largely heterogeneous: most items are distinct regressions or model-specific bugs, not duplicates. I would not merge any of the soft-edge pairs.", - "confidence": 0.94, - "canonical_issue_reason": "No single canonical issue stands out; the set spans unrelated bugs across loading, configs, processors, models, and docs.", + "summary": "This cluster is very noisy: the items span unrelated bugs in tokenizers, model loading, vision/audio processors, configs, docs, and backend availability. None of the soft pairs look like true duplicates or the same fixable code-path issue.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43873 is the broadest, highest-engagement runtime bug in the set, so it is the best single representative even though the overall cluster is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "No issue cleanly subsumes the whole set. If a placeholder representative is needed, 43749 is the closest because it is the most discussed broad runtime/loading regression, but it still does not represent the rest well.", + "best_issue_reason": "issue:43873 has the most discussion/inbound refs and the widest apparent user impact, making it the most suitable anchor issue for this noisy group.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44514", - "right": "issue:45381", + "left": "issue:43066", + "right": "issue:44462", "accept": false, - "reason": "Both involve Qwen2.5-VL video input, but one is a batching/padding crash and the other is a vision_position_ids correctness bug; not the same fix." + "reason": "Both concern tokenizer behavior, but one is a decoder-type regression and the other is repository tokenizer file selection; different bugs and code paths." }, { - "left": "issue:46121", - "right": "issue:46399", + "left": "issue:45588", + "right": "issue:46821", "accept": false, - "reason": "Rope-parameter JSON typing and kernel registration/decorator failure are different code paths and different failure modes." + "reason": "Unrelated failures: flash_attention sink handling vs TimesFM window_size attribute access." }, { - "left": "issue:43262", - "right": "issue:44206", + "left": "issue:29870", + "right": "issue:29925", "accept": false, - "reason": "Audio chat-template sampling-rate default vs feature-extractor center-arg crash are unrelated bugs." + "reason": "Bart generate() decoder_attention_mask bug and SigLIP fast-tokenizer enforcement are separate tokenizer/generation issues." }, { - "left": "issue:43761", - "right": "issue:45859", + "left": "issue:25251", + "right": "issue:28282", "accept": false, - "reason": "CLIPVision hidden_states regression and Qwen3.5 MoE CPU-offload incompatibility are unrelated model-specific issues." + "reason": "Pipeline top_k nesting bug is unrelated to missing PyTorch ImportError during AutoModel loading." }, { - "left": "issue:43316", - "right": "issue:47312", + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Different model families and failures: FastSpeech2Conformer tokenizer load vs Perceiver image-size interpolation." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to LayoutLMv3 box validation messaging." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL, but one is batched chat-template padding and the other is video vision_position_ids; not the same bug." + }, + { + "left": "issue:46121", + "right": "issue:46399", "accept": false, - "reason": "Gemma3TextConfig API mismatch and kernels 0.16 incompatibility are not the same underlying bug." + "reason": "Rope-parameter JSON/list handling is unrelated to Qwen3.5 kernel registration/decorator failure." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIP hidden_states regression and Qwen3.5 MoE rotary embedding CPU offload incompatibility are distinct issues." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "Broad custom-model-init breakage and patchtsmixer post_init policy are different problems, even if both touch initialization." + "reason": "Generic custom model init breakage is not the same as patchtsmixer's post_init API constraint." }, { "left": "issue:46821", "right": "issue:47436", "accept": false, - "reason": "TimesFM window_size attribute handling and Qwen2_5OmniConfig missing max_position_embeddings are unrelated config errors." + "reason": "TimesFM window_size attribute error and Qwen2_5OmniConfig missing max_position_embeddings are unrelated config/loading bugs." }, { - "left": "issue:43720", - "right": "issue:44442", + "left": "issue:43316", + "right": "issue:47312", "accept": false, - "reason": "BitNet accelerated-loading packed-weight failure and tokenizer load failure are unrelated." + "reason": "Gemma3TextConfig API mismatch is unrelated to kernels 0.16 incompatibility." }, { - "left": "issue:25251", - "right": "issue:28282", + "left": "issue:43720", + "right": "issue:44442", "accept": false, - "reason": "Pipeline top_k output nesting and missing PyTorch import are completely different issues." + "reason": "BitNet accelerate-loading unpacking bug and FastSpeech2Conformer tokenizer load failure are unrelated." }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Recurrent-state reset in modular_qwen3_5 and Whisper batch_decode skip_special_tokens are unrelated behaviors." + "reason": "Qwen recurrent-state reset bug and Whisper batch_decode skip_special_tokens behavior affect different code paths." }, { - "left": "issue:44977", - "right": "issue:45245", + "left": "issue:42175", + "right": "issue:46234", "accept": false, - "reason": "Qwen3.5 flash-attention generation bug and category-cardinality limit error do not share a concrete code path." + "reason": "Backend packaging issue with pip extras is unrelated to a missing import_utils availability helper." }, { - "left": "issue:42175", - "right": "issue:46234", + "left": "issue:43262", + "right": "issue:44206", "accept": false, - "reason": "Packaging/backend dependency mismatch and missing import_utils helper are different issues." + "reason": "Audio chat-template sample-rate defaulting and LasrFeatureExtractor center-arg crash are separate audio bugs." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "A model-level post_init restriction is not the same as the Qwen video-input regression." + "reason": "PatchTSMixer post_init restriction is unrelated to Qwen video-input regression." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading behavior and Gemma eos_token_id inconsistency are unrelated." + "reason": "Quantization/offloading failure and Gemma eos_token_id mismatch are different runtime/config issues." }, { "left": "issue:43316", "right": "issue:45759", "accept": false, - "reason": "Both are config-related, but one is a Gemma API discrepancy and the other is AutoModelForCausalLM unwrapping text_config for Qwen multimodal configs." + "reason": "Both involve model config APIs, but Gemma3TextConfig discrepancy and Qwen composite text_config unwrapping are different defects." }, { "left": "issue:42490", "right": "issue:44743", "accept": false, - "reason": "Local-path trailing-slash import warning and recurrent-state reset are unrelated." + "reason": "Local-path import warning has no relation to Qwen recurrent-state reset behavior." }, { - "left": "issue:43493", - "right": "issue:43827", + "left": "issue:44977", + "right": "issue:45245", "accept": false, - "reason": "SigLIP2 implementation divergence and docs still referencing pipeline() are not the same bug." + "reason": "Flash-attention generation failure and category-count limit error are unrelated." }, { "left": "issue:47276", "right": "issue:47448", "accept": false, - "reason": "A misleading warning message and Gemma4 architecture recognition are unrelated." + "reason": "Misleading synced_gpus warning and Gemma4 arch recognition are different issues." }, { - "left": "issue:43749", - "right": "issue:43931", + "left": "issue:43493", + "right": "issue:43827", "accept": false, - "reason": "Both are loading problems, but FSDP CPU-efficient loading and Qwen3-VL weight-shape mismatch are different failures." + "reason": "SigLIP2 implementation discrepancy and docs still referencing pipeline() are unrelated." }, { - "left": "issue:43643", - "right": "issue:44913", + "left": "issue:43749", + "right": "issue:43931", "accept": false, - "reason": "Missing fields with trust_remote_code and GPTNeoX rotary_pct reload behavior are different config-loading issues." + "reason": "FSDP CPU-efficient loading breakage and Qwen3-VL weight-shape mismatch are different loading problems." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py execution failure and tensor-parallel OOM are not the same underlying bug." - }, - { - "left": "issue:44987", - "right": "issue:46178", - "accept": false, - "reason": "Transformers>=5.1.0 loading failure for one model and tensorflow_text backend mapping error are unrelated." + "reason": "3d_parallel.py runtime failure and Qwen2-57B-A14B-Instruct TP OOM are not the same bug." }, { - "left": "issue:43550", - "right": "issue:43701", - "accept": false, - "reason": "torch.compile/SDPA failure and resume_from_checkpoint key mismatch are different code paths." - }, - { - "left": "issue:44964", - "right": "issue:45440", - "accept": false, - "reason": "Phi-4 multimodal load failure and DeepseekV3MoE divergence are unrelated model-specific regressions." - }, - { - "left": "issue:43316", - "right": "issue:43525", - "accept": false, - "reason": "Both are config attribute issues, but Gemma3TextConfig API discrepancy and Llama4Config pad_token_id missing are not the same bug." - }, - { - "left": "issue:27957", - "right": "issue:29870", + "left": "issue:43643", + "right": "issue:44913", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta and BART generate UnboundLocalError are unrelated." + "reason": "Missing fields from trust_remote_code AutoConfig and GPTNeoX rotary_pct reset on reload are separate config persistence bugs." } ] }, @@ -11934,15 +11994,16 @@ "issue:23647", "issue:27957", "issue:28218", + "issue:29870", "issue:32090", "issue:43232", "issue:43295", "issue:43316", "issue:43388", - "issue:43452", + "issue:43525", + "issue:43550", "issue:43575", "issue:43606", - "issue:43646", "issue:43701", "issue:43716", "issue:43742", @@ -11961,15 +12022,15 @@ "issue:44534", "issue:44855", "issue:44871", - "issue:44913", + "issue:44964", + "issue:44987", "issue:45137", "issue:45216", + "issue:45440", "issue:45584", "issue:45803", "issue:45907", - "issue:46050", - "issue:46234", - "issue:46291", + "issue:46178", "issue:46506", "issue:46829", "issue:46932", @@ -11978,211 +12039,224 @@ "issue:47448" ], "soft_pairs": [ - "issue:43825|issue:47448", + "issue:44987|issue:46178", + "issue:43550|issue:43701", + "issue:43316|issue:43525", + "issue:43950|issue:47427", + "issue:27957|issue:29870", + "issue:44964|issue:45440", "issue:43606|issue:43701", "issue:43575|issue:43742", - "issue:43950|issue:47427", - "issue:44062|issue:45907", + "issue:43825|issue:47448", "issue:43761|issue:44079", + "issue:44062|issue:45907", + "issue:43316|issue:43716", "issue:44855|issue:45137", "issue:43701|issue:46506", - "issue:44190|issue:44496", - "issue:43316|issue:43716", "issue:46829|issue:46932", - "issue:44442|issue:45803", + "issue:44190|issue:44496", "issue:43931|issue:45216", + "issue:44442|issue:45803", "issue:27957|issue:32090", "issue:23647|issue:28218", "issue:43295|issue:44871", "issue:44871|issue:45584", "issue:43761|issue:44188", "issue:43232|issue:43388", - "issue:16998|issue:22355", - "issue:44521|issue:47217", - "issue:46234|issue:46291", "issue:44077|issue:44534", - "issue:43452|issue:44496", - "issue:43646|issue:44913", - "issue:44871|issue:46050" + "issue:16998|issue:22355", + "issue:44521|issue:47217" ], - "prepared_review_unit_hash": "e922f05cba9c2bb296d2f1367b09db512f5665f3d099cfede976a5629dea9b08", + "prepared_review_unit_hash": "d78df4f152a4036e683f7a4807ae571210729a86599a55a69653b574fad3f335", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 26, - "serialized_chars": 34924, - "estimated_input_tokens": 8731, - "estimated_eval_tokens": 17718 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35850, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items do not form a true duplicate cluster: they span unrelated loading, tokenization, generation, config, and training issues. I rejected all soft edges and found no single canonical duplicate hub.", - "confidence": 0.79, - "canonical_issue_reason": "No strong canonical issue: the cluster is too heterogeneous for a single shared bug. If forced to pick a representative, issue 44521 is the closest hub only because it is open and has the most inbound references/discussion, not because it duplicates the others.", + "summary": "The soft pairs are mostly false positives: they share broad Transformer themes but describe different code paths, components, or regression types. This is not a coherent duplicate cluster.", + "confidence": 0.96, + "canonical_issue_reason": "No solid canonical issue: the set mixes unrelated bugs across tokenizers, configs, generation, loading, training, and model export/runtime behavior.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44521 is the best representative of the set only in the weak sense that it is active, open, and most discussed; however, it should not be treated as the canonical duplicate target for the whole cluster.", + "best_issue_reason": "issue:44521 is the most active/open item with the most inbound references, so it is the best available cluster anchor, but it still does not unify the rest of the issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43825", - "right": "issue:47448", + "left": "issue:44987", + "right": "issue:46178", "accept": false, - "reason": "Translation-task error messaging and Gemma4 architecture recognition are unrelated bugs in different code paths." + "reason": "Different failures: model loading for a specific repo vs backend registration for tensorflow_text." }, { - "left": "issue:43606", + "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch and resume_from_checkpoint key mismatch are different training/runtime failures." + "reason": "Unrelated problems: torch.compile SDPA bug vs checkpoint key mismatch on resume." }, { - "left": "issue:43575", - "right": "issue:43742", + "left": "issue:43316", + "right": "issue:43525", "accept": false, - "reason": "Tensor-parallel OOM on Qwen2-57B and a key error loading MobileLLM are distinct load-time problems." + "reason": "Both are config/API issues, but one is Gemma3TextConfig discrepancy and the other is missing pad_token_id on Llama4Config." }, { "left": "issue:43950", "right": "issue:47427", "accept": false, - "reason": "Both touch initialization/loading, but one corrupts non-persistent buffers while the other skips redundant init checks; not the same bug." + "reason": "Different code paths: corrupted non-persistent buffers on load vs redundant HF init checks on non-CUDA hardware." }, { - "left": "issue:44062", - "right": "issue:45907", + "left": "issue:27957", + "right": "issue:29870", "accept": false, - "reason": "AddedToken special-argument duplication and list_repo_files revision forwarding are unrelated issues." + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Bart generate() using decoder_attention_mask before assignment." }, { - "left": "issue:43761", - "right": "issue:44079", + "left": "issue:44964", + "right": "issue:45440", "accept": false, - "reason": "A CLIPVisionModel hidden_states regression and a generic ModelOutput None-key assignment bug are different code paths." + "reason": "Different model families and bugs: Phi-4 multimodal load failure vs DeepSeekV3MoE implementation divergence." }, { - "left": "issue:44855", - "right": "issue:45137", + "left": "issue:43606", + "right": "issue:43701", "accept": false, - "reason": "Python 3.13 import-time IndentationError in DebertaV2Model is unrelated to a DeepSpeed ZeRO3 deque IndexError." + "reason": "CPU offload device mismatch is unrelated to resume_from_checkpoint key mismatch." }, { - "left": "issue:43701", - "right": "issue:46506", + "left": "issue:43575", + "right": "issue:43742", "accept": false, - "reason": "Checkpoint resume key mismatch and CPU/CUDA output divergence are not the same underlying defect." + "reason": "Qwen2 tensor-parallel OOM and MobileLLM config key error are different loading issues." }, { - "left": "issue:44190", - "right": "issue:44496", + "left": "issue:43825", + "right": "issue:47448", + "accept": false, + "reason": "One is an error-message regression in pipeline(), the other is gemma4 architecture recognition." + }, + { + "left": "issue:43761", + "right": "issue:44079", "accept": false, - "reason": "Local dataset loading in a training script and unrecognized-model config loading are separate problems." + "reason": "CLIPVisionModel hidden_states output bug is different from ModelOutput key assignment when previous value is None." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer AddedToken argument collision is unrelated to list_repo_files() not forwarding revision." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Gemma3TextConfig API discrepancy and Mistral-3 image-preprocessor dtype mismatch are unrelated." + "reason": "Gemma3TextConfig API mismatch and Mistral-3 vision dtype mismatch are separate regressions." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 indentation/import parsing issue is unrelated to DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "Checkpoint resume key mismatch and CPU/CUDA output divergence are different problems." }, { "left": "issue:46829", "right": "issue:46932", "accept": false, - "reason": "VideoPrism hidden_states misassignment and Gemma4 quantized-embedder dtype casting are different bugs." + "reason": "VideoPrism hidden_states misuse and Gemma4 quantized embedder dtype casting do not look like the same bug." }, { - "left": "issue:44442", - "right": "issue:45803", + "left": "issue:44190", + "right": "issue:44496", "accept": false, - "reason": "AutoTokenizer loading failure and a bare-except tensor conversion bug do not describe the same issue." + "reason": "Local dataset loading in a training script is unrelated to unrecognized model/config loading." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch on load and Qwen3.5 checkpoint corruption on save are related only by model family, not by the same bug." + "reason": "One is a shape mismatch when loading a model, the other is a save_pretrained regression." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "AutoTokenizer loading failure and a bare except swallowing errors in FuyuBatchFeature are different defects." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "Flash Attention support for XLMRoberta and a Trainer NoneType broadcast error are unrelated." + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Trainer GPU broadcast failing on NoneType." }, { "left": "issue:23647", "right": "issue:28218", "accept": false, - "reason": "Wandb sweeps custom objective support and tokenizer spacing after added tokens are unrelated features/bugs." + "reason": "Wandb sweeps custom objective and tokenizer added-space behavior are unrelated." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "Processor/tokenizer access regression and Gemma3 eos_token_id inconsistency are different model/config issues." + "reason": "Processor/tokenizer access regression is different from Gemma eos_token_id configuration mismatch." }, { "left": "issue:44871", "right": "issue:45584", "accept": false, - "reason": "Gemma3 eos_token_id mismatch and Whisper empty-transcription generation failure are not the same bug." + "reason": "Gemma eos_token_id inconsistency and Whisper empty transcription after align_special_tokens are distinct bugs." }, { "left": "issue:43761", "right": "issue:44188", "accept": false, - "reason": "CLIPVisionModel hidden_states regression and torch.compile attention-kernel divergence are separate defects." + "reason": "Hidden_states regression and attention-kernel divergence on torch.compile are unrelated." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation kwargs sync issue and gather_for_metrics tuple-label truncation are unrelated distributed-training bugs." - }, - { - "left": "issue:16998", - "right": "issue:22355", - "accept": false, - "reason": "A model_max_length question for DeBERTa-V3 and a missing transformers.onnx import are unrelated." - }, - { - "left": "issue:44521", - "right": "issue:47217", - "accept": false, - "reason": "Both are multimodal/chat-template adjacent, but zeroed assistant masks and failure when the image placeholder appears in user text are different underlying problems." - }, - { - "left": "issue:46234", - "right": "issue:46291", - "accept": false, - "reason": "Missing import_utils availability helper and pipelines import failing on kernels version/revision requirements are distinct dependency issues." + "reason": "Generation kwargs update after sync_gpus and gather_for_metrics label truncation are different training/generation paths." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "PatchTSMixer post_init policy and non-persistent buffer corruption in v5 are unrelated regressions." + "reason": "PatchTSMixer post_init policy and non-persistent buffer corruption are not the same issue." }, { - "left": "issue:43452", - "right": "issue:44496", + "left": "issue:16998", + "right": "issue:22355", "accept": false, - "reason": "gguf_file breaking from_pretrained and unrecognized-model config.json errors are different loading paths and failure modes." + "reason": "model_max_length question on DeBERTa-V3 and missing transformers.onnx module are unrelated." }, { - "left": "issue:43646", - "right": "issue:44913", + "left": "issue:44521", + "right": "issue:47217", "accept": false, - "reason": "A broad Transformers 5 custom initialization regression and GPTNeoXConfig rotary_pct reload defaulting are not the same code-path bug." + "reason": "Both touch multimodal chat/template handling, but all-zero assistant masks and image-placeholder-in-user-text failures are different underlying bugs." } ] }, @@ -12217,31 +12291,33 @@ "issue:43352", "issue:43404", "issue:43441", + "issue:43452", "issue:43504", "issue:43525", "issue:43531", "issue:43540", + "issue:43646", "issue:43931", "issue:44060", "issue:44162", "issue:44464", "issue:44479", + "issue:44496", "issue:44514", "issue:44534", "issue:44610", "issue:44617", "issue:44792", + "issue:44871", "issue:44913", "issue:44977", - "issue:45127", - "issue:45200", "issue:45381", + "issue:45412", "issue:45542", - "issue:45588", "issue:45600", - "issue:45632", "issue:45698", "issue:45803", + "issue:46050", "issue:46170", "issue:46178", "issue:46234", @@ -12249,238 +12325,250 @@ "issue:46365", "issue:46369", "issue:46421", - "issue:46424", "issue:46640", "issue:46693", - "issue:47013", "issue:47217" ], "soft_pairs": [ - "issue:44464|issue:46421", + "issue:46234|issue:46291", + "issue:43452|issue:44496", "issue:25251|issue:27957", + "issue:43646|issue:44913", + "issue:44871|issue:46050", + "issue:43352|issue:46640", + "issue:45412|issue:46369", "issue:43404|issue:43441", + "issue:44464|issue:46421", "issue:43065|issue:44617", - "issue:42907|issue:45698", - "issue:43352|issue:46640", - "issue:44162|issue:44792", "issue:43540|issue:44514", + "issue:44162|issue:44792", + "issue:42907|issue:45698", "issue:46178|issue:46291", "issue:44792|issue:44977", - "issue:44610|issue:46369", "issue:44479|issue:47217", - "issue:45803|issue:46365", + "issue:44610|issue:46369", "issue:45381|issue:47217", - "issue:45542|issue:46234", "issue:44534|issue:44913", - "issue:44792|issue:45600", + "issue:45542|issue:46234", "issue:43525|issue:43531", + "issue:45803|issue:46365", + "issue:44792|issue:45600", "issue:43504|issue:45698", "issue:27957|issue:29925", "issue:46693|issue:47217", - "issue:45803|issue:46170", - "issue:23647|issue:27957", "issue:35141|issue:42490", - "issue:45632|issue:46424", - "issue:43931|issue:44060", - "issue:45127|issue:45200", - "issue:45588|issue:47013" + "issue:23647|issue:27957", + "issue:45803|issue:46170", + "issue:43931|issue:44060" ], - "prepared_review_unit_hash": "70277ec307d1c8ac970272f7530507c8922be8037d88077f12beabe60f425bf0", + "prepared_review_unit_hash": "b6b626aabf7c11ac7c414b1819cf865edfc7186a2bdb2624d45a35906beaf013", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 35992, - "estimated_input_tokens": 8998, - "estimated_eval_tokens": 18252 + "soft_pair_count": 30, + "serialized_chars": 36531, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly noise: the issues span many unrelated models and subsystems (Flash Attention, tokenization, cache, multimodal processors, config/loading bugs, etc.), so they should not be merged as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue fits the cluster; the items are disparate and describe different underlying bugs rather than one repeated defect.", + "summary": "The cluster is thematically broad and noisy: most items are separate Transformers issues around model loading, config serialization, multimodal inputs, or Flash Attention behavior. None of the suggested soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45698 is the best representative: it is concrete, well-described, and has strong discussion activity around a save/load custom-module failure, which fits the main loading/serialization theme in this cluster.", "canonical_pr_reason": null, - "best_issue_reason": "No clear best issue either, because the cluster is not centered on one concrete problem. If forced, none of these is representative of the others.", + "best_issue_reason": "issue:45698 is the most suitable single issue to anchor the cluster because it is specific, reproducible, and broadly representative of the save/load regressions appearing here.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44464", - "right": "issue:46421", + "left": "issue:46234", + "right": "issue:46291", "accept": false, - "reason": "Both mention chunked generation/cache behavior, but one is about inconsistent outputs with compiled forward and the other about spurious recompilation from cache lazy init; different bugs." + "reason": "Different failures: missing utility symbol vs pipeline import dependency check." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Both are loading errors, but one is gguf_file handling and the other is an unrecognized model config; not the same bug." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "Top-k pipeline nesting and XLM-RoBERTa Flash Attention 2 are unrelated features and failure modes." + "reason": "Unrelated: pipeline top_k output shape vs XLM-RoBERTa Flash Attention 2 support." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Different regressions: custom model init under v5 vs GPTNeoX config reload losing rotary_pct." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Both involve model settings after load, but one is eos_token_id mismatch and the other is quantized sinq save/load breakage." + }, + { + "left": "issue:43352", + "right": "issue:46640", + "accept": false, + "reason": "Specific unsupported model error vs generic Flash Attention 2 issue; not the same code-path." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "One is GPU memory not released after deletion, the other is runtime OOM during Sam3-lite text execution." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "One is a tied-weights bug in Mistral3ForConditionalGeneration; the other is a Flash Attention support error for Ministral-3. Same family, different defects." + "reason": "Different Mistral/Ministral problems: tied lm_head weights vs FlashAttention failure." }, { - "left": "issue:43065", - "right": "issue:44617", + "left": "issue:44464", + "right": "issue:46421", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same underlying problem as Sam3Video OOM." + "reason": "Both touch generation/cache behavior, but one is incorrect outputs and the other is spurious recompilation from lazy cache init." }, { - "left": "issue:42907", - "right": "issue:45698", + "left": "issue:43065", + "right": "issue:44617", "accept": false, - "reason": "Saving dequantized models and loading the wrong custom module after save_pretrained are different save/load issues." + "reason": "Same model family, but dummy Conv2d vs OOM are different defects." }, { - "left": "issue:43352", - "right": "issue:46640", + "left": "issue:43540", + "right": "issue:44514", "accept": false, - "reason": "A model-specific FA2 unsupported error is not the same as a generic Flash Attention 2 issue." + "reason": "Different multimodal bugs: video processing in Qwen3OmniMoe vs batched chat-template padding crash in Qwen2.5-VL." }, { "left": "issue:44162", "right": "issue:44792", "accept": false, - "reason": "ESM2 breakage and a janus image-generation test failure are unrelated." + "reason": "ESM2 breakage is unrelated to a janus image-generation test failure." }, { - "left": "issue:43540", - "right": "issue:44514", + "left": "issue:42907", + "right": "issue:45698", "accept": false, - "reason": "Both are Qwen multimodal bugs, but one is video-input processing in Qwen3OmniMoe and the other is batched chat-template padding behavior in Qwen2_5_VLProcessor." + "reason": "Both are from_pretrained/save_pretrained-adjacent, but one is a trailing-slash path warning and the other loads the wrong custom module." }, { "left": "issue:46178", "right": "issue:46291", "accept": false, - "reason": "Backend mapping for tensorflow_text and kernels import-version requirements are different import-time dependency problems." + "reason": "Different import failures: tensorflow_text backend mapping vs kernels package version/revision requirement." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus generation test failure and Qwen3.5 flash-attention generation issues are not the same defect." - }, - { - "left": "issue:44610", - "right": "issue:46369", - "accept": false, - "reason": "Processor image size mismatch and Sam3-lite text memory overflow are separate problems." + "reason": "Janus image-generation test failure is unrelated to Qwen3.5 Flash Attention generation behavior." }, { "left": "issue:44479", "right": "issue:47217", "accept": false, - "reason": "A video-input regression in Qwen VL models is not the same as image-placeholder handling in user text." + "reason": "Both are VL regressions, but one is video-input handling after v5.3.0 and the other is image-placeholder placement in user text." }, { - "left": "issue:45803", - "right": "issue:46365", + "left": "issue:44610", + "right": "issue:46369", "accept": false, - "reason": "Bare except swallowing errors in Fuyu tensors conversion is unrelated to Trainer import failure from datasets." + "reason": "Different resource/input bugs: processor output size mismatch vs memory overflow in Sam3-lite text." }, { "left": "issue:45381", "right": "issue:47217", "accept": false, - "reason": "Both involve VL prompting, but wrong vision_position_ids for video input is different from failure when the image placeholder appears in user text." + "reason": "Both involve VL inputs, but one is wrong vision_position_ids for video input and the other is image placeholder handling." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk in v5 is unrelated to GPTNeoX rotary_pct reload behavior." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "Tensorboard-without-TensorFlow backend errors and a missing import_utils helper are unrelated." + "reason": "TensorBoard/TensorFlow backend issue is unrelated to a missing falcon_mamba utility function." }, { - "left": "issue:44534", - "right": "issue:44913", + "left": "issue:43525", + "right": "issue:43531", "accept": false, - "reason": "Non-persistent buffer initialization junk in v5 is not the same as GPTNeoXConfig rotary_pct not persisting on reload." + "reason": "Different models and failures: missing pad_token_id vs sliding_window issue." }, { - "left": "issue:44792", - "right": "issue:45600", + "left": "issue:45803", + "right": "issue:46365", "accept": false, - "reason": "A janus image-generation test failure and stale auto_mappings references to removed Sam3LiteText configs are different issues." + "reason": "Bare except swallowing errors in Fuyu conversion is unrelated to Trainer import conflicts with datasets." }, { - "left": "issue:43525", - "right": "issue:43531", + "left": "issue:44792", + "right": "issue:45600", "accept": false, - "reason": "Missing pad_token_id on Llama4Config and Qwen3-MoE sliding_window behavior are unrelated config/model bugs." + "reason": "A failed janus test is not the same as stale auto_mappings references to removed Sam3LiteText configs." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "Legacy-field loading failure in BEiT segmentation and wrong custom-module loading after save_pretrained are different load-path bugs." + "reason": "Different loading bugs: legacy field preset loading vs wrong custom-module restoration after save_pretrained." }, { "left": "issue:27957", "right": "issue:29925", "accept": false, - "reason": "Flash Attention 2 support for XLM-RoBERTa and SigLIP fast-tokenizer enforcement are unrelated." + "reason": "Flash Attention 2 support for XLM-RoBERTa is unrelated to SigLIP tokenizer fast-mode enforcement." }, { "left": "issue:46693", "right": "issue:47217", "accept": false, - "reason": "FA2 long-context performance regression is not the same as VL image-placeholder placement failure." + "reason": "Different VL/performance problems: FA2 long-context regression vs image-placeholder failure." }, { - "left": "issue:45803", - "right": "issue:46170", + "left": "issue:35141", + "right": "issue:42490", "accept": false, - "reason": "Fuyu tensor conversion error handling and Windows ROCm/torch.distributed import issues are unrelated." + "reason": "Embedding reinit on resize is unrelated to trailing-slash local path import warnings." }, { "left": "issue:23647", "right": "issue:27957", "accept": false, - "reason": "Wandb sweep custom objective integration has nothing to do with XLM-RoBERTa Flash Attention 2." - }, - { - "left": "issue:35141", - "right": "issue:42490", - "accept": false, - "reason": "Embedding reinitialization on resize and local-path trailing-slash import warnings are different loading/configuration problems." + "reason": "W&B sweeps custom objective is unrelated to XLM-RoBERTa Flash Attention 2." }, { - "left": "issue:45632", - "right": "issue:46424", + "left": "issue:45803", + "right": "issue:46170", "accept": false, - "reason": "Cache-path collisions for remote code and max_cache_length handling in quantized cache are separate cache issues." + "reason": "Different platforms and bugs: Fuyu exception handling vs Windows ROCm import dependency problems." }, { "left": "issue:43931", "right": "issue:44060", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch and Qwen3-Next tied-weights warning are different model bugs." - }, - { - "left": "issue:45127", - "right": "issue:45200", - "accept": false, - "reason": "LoRA merge collapse with extended vocab and Gemma 4 mm_token_type_ids defaults are unrelated." - }, - { - "left": "issue:45588", - "right": "issue:47013", - "accept": false, - "reason": "Flash-attention sink-less model crash and xLSTM train-mode unpacking failure are different code paths." + "reason": "Both are Qwen3-family loading/warning issues, but one is a weight-shape mismatch and the other is a tied-weights warning bug; not the same defect." } ] }, @@ -12509,13 +12597,10 @@ "issue:29127", "issue:29870", "issue:35141", - "issue:39692", - "issue:41720", "issue:41950", "issue:43335", "issue:43388", "issue:43475", - "issue:43550", "issue:43644", "issue:43650", "issue:43673", @@ -12534,251 +12619,240 @@ "issue:44913", "issue:44964", "issue:45072", + "issue:45127", + "issue:45200", "issue:45245", "issue:45381", "issue:45405", "issue:45406", - "issue:45563", "issue:45584", "issue:45588", + "issue:45632", "issue:46082", "issue:46202", "issue:46369", "issue:46399", + "issue:46424", "issue:46697", "issue:46832", "issue:46858", "issue:46894", "issue:46932", - "issue:47312", - "issue:47321" + "issue:47013", + "issue:47312" ], "soft_pairs": [ - "issue:43673|issue:44743", + "issue:45632|issue:46424", + "issue:45127|issue:45200", + "issue:25251|issue:29870", + "issue:45588|issue:47013", "issue:44855|issue:45588", + "issue:43673|issue:44743", "issue:43650|issue:46202", "issue:41950|issue:43475", "issue:44242|issue:46894", "issue:46082|issue:46697", "issue:44792|issue:45405", - "issue:44190|issue:44964", - "issue:43388|issue:43688", - "issue:43827|issue:43994", "issue:44617|issue:45584", - "issue:45245|issue:46858", - "issue:25251|issue:29870", + "issue:43827|issue:43994", + "issue:43388|issue:43688", + "issue:44190|issue:44964", "issue:45072|issue:46932", "issue:29870|issue:35141", "issue:43644|issue:44913", "issue:44734|issue:45406", "issue:44869|issue:46832", - "issue:43335|issue:44855", - "issue:46399|issue:47312", + "issue:45245|issue:46858", "issue:23647|issue:29127", + "issue:43335|issue:44855", "issue:43854|issue:44792", - "issue:45381|issue:46369", - "issue:41720|issue:45245", - "issue:39692|issue:47321", - "issue:43550|issue:44190", - "issue:45563|issue:45584", - "issue:43827|issue:47321" + "issue:46399|issue:47312", + "issue:45381|issue:46369" ], - "prepared_review_unit_hash": "820e82093f12f73dbf701215ce43582afddd74db1aed8a56ca0cf0423f0deeb4", + "prepared_review_unit_hash": "b7b2923cfd5707c7eca187a0a7053d6523c416e5df98ca19fb5ca3fb7c166b42", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36263, - "estimated_input_tokens": 9066, - "estimated_eval_tokens": 18388 + "soft_pair_count": 26, + "serialized_chars": 35833, + "estimated_input_tokens": 8959, + "estimated_eval_tokens": 18174 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is very heterogeneous: most items are unrelated bugs across different models, pipelines, and internals. No soft pair looks like the same underlying issue, so I would not merge any of them as duplicates.", - "confidence": 0.19, - "canonical_issue_reason": "No true duplicate cluster is evident; issue #43644 is the broadest, most central core-regression report among the set, so it is the best placeholder canonical issue.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers issues across different models, pipelines, and runtime paths. None of the soft pairs look like true duplicates or mergeable as one fix; at best they share broad keywords like cache, dtype, or serving.", + "confidence": 0.84, + "canonical_issue_reason": "If a representative issue must be chosen, issue:43644 is the cleanest broad regression report: it has a concrete v5.0 behavior change, a core code-path impact, and a clear problem statement. That said, the cluster does not really have a single canonical duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "#43644 is the strongest representative because it describes a general runtime regression in core Transformers behavior and has more cluster significance than the narrowly scoped model-specific reports.", + "best_issue_reason": "issue:43644 is the strongest standalone issue in the set because it describes a specific, high-impact regression with a general scope and a likely actionable fix path.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43673", - "right": "issue:44743", + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache, but one is a local trust_remote_code path collision and the other is cache max length handling for quantized cache; different bugs and code paths." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "Extended-vocab LoRA collapse vs mm_token_type_ids defaults for Gemma 4 text-only fine-tuning are unrelated model/config bugs." + }, + { + "left": "issue:25251", + "right": "issue:29870", + "accept": false, + "reason": "Top-k pipeline nesting change and BART generate decoder_attention_mask initialization failure are different subsystems and failure modes." + }, + { + "left": "issue:45588", + "right": "issue:47013", "accept": false, - "reason": "Both involve cache/state behavior in generation, but one is a v5 chunked_prefill cache issue and the other is recurrent-state reset in modular Qwen3.5; different code paths and fixes." + "reason": "Flash-attention s_aux=None crash and xLSTM return_last_states unpacking bug affect different layers and code paths; not the same fix." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 import parsing failure in DebertaV2 is unrelated to flash_attention handling of s_aux=None." + "reason": "Python 3.13 IndentationError in DebertaV2Model import is unrelated to a flash-attention AttributeError." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Both involve cache, but one is missing GenerationMixin cache in chunked prefill while the other is recurrent-state reset in modular_qwen3_5; distinct defects." }, { "left": "issue:43650", "right": "issue:46202", "accept": false, - "reason": "'ADD THE DATA' is unrelated to Gemma pipeline roundtrip failures on the underscore token." + "reason": "A placeholder issue and a Gemma pipeline roundtrip bug are clearly unrelated." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Both are vision-related, but one is a pipeline processor lookup bug and the other is a missing attribute in Sam3 vision encoder output." + "reason": "Video-classification processor selection and SAM3 vision encoder missing attribute are different vision-stack bugs." }, { "left": "issue:44242", "right": "issue:46894", "accept": false, - "reason": "Both concern training losses, but they are different bugs in different model families and loss computations; not the same concrete fix." + "reason": "Load-balancing loss gating and ASR training-loss double-shift are both training-loss issues, but they are separate implementation errors." }, { "left": "issue:46082", "right": "issue:46697", "accept": false, - "reason": "LlamaConfig head_dim validation and Swin2SR image over-padding are unrelated." + "reason": "Llama head_dim validation and Swin2SR over-padding are unrelated config vs preprocessing issues." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "A failed Janus image generation test is unrelated to a PEFT minimum-version bump issue." - }, - { - "left": "issue:44190", - "right": "issue:44964", - "accept": false, - "reason": "Dataset loading in a trainer script and Phi-4 multimodal model loading are different failure modes." + "reason": "A failing janus generate-images test and an unreleased MIN_PEFT_VERSION bump do not share the same underlying bug." }, { - "left": "issue:43388", - "right": "issue:43688", + "left": "issue:44617", + "right": "issue:45584", "accept": false, - "reason": "Metric gathering label truncation and auxiliary-loss normalization are unrelated internals." + "reason": "SAM3 video OOM and Whisper empty-transcription generation failure are unrelated model/runtime problems." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs still referencing pipeline() is not the same as a SigLIP2 model/pipeline correctness problem." + "reason": "Docs still referencing pipeline() and SIGLIP2 producing bad results with AutoModel/pipeline are different issues; one is documentation, the other runtime behavior." }, { - "left": "issue:44617", - "right": "issue:45584", - "accept": false, - "reason": "CUDA OOM in Sam3Video is unrelated to Whisper empty-transcription decoding failure." - }, - { - "left": "issue:45245", - "right": "issue:46858", + "left": "issue:43388", + "right": "issue:43688", "accept": false, - "reason": "Category-count limit and static-cache device-tensor arithmetic are distinct bugs." + "reason": "Metric-gathering truncation and auxiliary-loss normalization are both training-related, but the failure mechanisms and affected code are different." }, { - "left": "issue:25251", - "right": "issue:29870", + "left": "issue:44190", + "right": "issue:44964", "accept": false, - "reason": "Pipeline output shape regression and BART generate() unbound local error are different code paths." + "reason": "Local dataset loading and Phi-4 multimodal model loading are not the same bug." }, { "left": "issue:45072", "right": "issue:46932", "accept": false, - "reason": "Both mention dtype/casting, but one is bfloat16 inference mismatches and the other is quantized embedders in Gemma4; not mergeable into one fix." + "reason": "Both mention dtype mismatches, but they affect different models and different quantization/inference paths; too broad to merge." }, { "left": "issue:29870", "right": "issue:35141", "accept": false, - "reason": "BART generate() decoder_attention_mask bug is unrelated to output-embedding reinitialization after resizing token embeddings." + "reason": "BART generate uses an uninitialized decoder_attention_mask; resizing embeddings reinitializes output embeddings in post_init. Different lifecycle bugs." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "A general non-persistent-buffer initialization issue is unrelated to GPTNeoXConfig rotary_pct not persisting on reload." + "reason": "Non-persistent buffers filled with junk and GPTNeoXConfig rotary_pct reload defaulting are unrelated state/config issues." }, { "left": "issue:44734", "right": "issue:45406", "accept": false, - "reason": "Both affect transformers serve, but one is KV-cache continuation indexing and the other is a missing processor attribute; different bugs." + "reason": "KV-cache continuation tensor indexing in serve and Gemma4Processor missing _tokenizer are separate serve-time failures." }, { "left": "issue:44869", "right": "issue:46832", "accept": false, - "reason": "Whisper timestamp decoding and RT-DETR feature-level handling are unrelated." - }, - { - "left": "issue:43335", - "right": "issue:44855", - "accept": false, - "reason": "SwitchTransformers sparse-layer construction and a DebertaV2 import syntax issue have nothing in common beyond being bugs." + "reason": "Whisper timestamp decoding on trailing replacement chars and RT-DETR forward feature-level mismatch are unrelated model-specific bugs." }, { - "left": "issue:46399", - "right": "issue:47312", + "left": "issue:45245", + "right": "issue:46858", "accept": false, - "reason": "Kernel registration decorator failure in Qwen3.5 is not clearly the same as a vague incompatibility with kernels 0.16." + "reason": "Category-count overflow and static-cache seq_length device-tensor arithmetic are different low-level errors." }, { "left": "issue:23647", "right": "issue:29127", "accept": false, - "reason": "Wandb sweeps custom objective integration and LayoutLMv3 error-message clarity are unrelated." - }, - { - "left": "issue:43854", - "right": "issue:44792", - "accept": false, - "reason": "GLM-4.7-Flash unit-test loading issues do not match a Janus image-generation test failure." - }, - { - "left": "issue:45381", - "right": "issue:46369", - "accept": false, - "reason": "Qwen2.5-VL video position IDs and Sam3-litetext memory overflow are different model/runtime problems." + "reason": "Wandb sweeps custom objective and LayoutLMv3 error-message clarity do not describe the same issue." }, { - "left": "issue:41720", - "right": "issue:45245", - "accept": false, - "reason": "Auto device mapping cuda assert and category-count overflow are unrelated." - }, - { - "left": "issue:39692", - "right": "issue:47321", + "left": "issue:43335", + "right": "issue:44855", "accept": false, - "reason": "SigLIP2 docs/model-processor issues are unrelated to AMD quark class updates." + "reason": "SwitchTransformers sparse-layer creation and a Python 3.13 import IndentationError are entirely different problems." }, { - "left": "issue:43550", - "right": "issue:44190", + "left": "issue:43854", + "right": "issue:44792", "accept": false, - "reason": "torch.compile SDPA failure in Bamba is unrelated to local dataset loading in an image-classification script." + "reason": "GLM-4.7-Flash loading in tests and janus image-generation test failure are unrelated model/test cases." }, { - "left": "issue:45563", - "right": "issue:45584", + "left": "issue:46399", + "right": "issue:47312", "accept": false, - "reason": "A stale generate() warning and Whisper empty-transcription failure are different severity and code paths." + "reason": "Missing decorator in Qwen3.5 rotary registration and kernel version incompatibility are both kernel-related, but not the same concrete bug." }, { - "left": "issue:43827", - "right": "issue:47321", + "left": "issue:45381", + "right": "issue:46369", "accept": false, - "reason": "Pipeline documentation cleanup has no relation to AMD quark class updates." + "reason": "Qwen2.5-VL vision_position_ids bug and SAM3 litetext memory overflow are separate model issues." } ] }, @@ -12803,22 +12877,21 @@ "cluster_id": "cluster-10000-544", "nodes": [ "issue:39692", + "issue:41720", "issue:43065", "issue:43295", "issue:43317", "issue:43335", "issue:43526", "issue:43540", - "issue:43611", + "issue:43550", "issue:43643", "issue:43673", + "issue:43827", "issue:43854", "issue:43994", - "issue:44016", - "issue:44112", "issue:44190", "issue:44242", - "issue:44279", "issue:44479", "issue:44496", "issue:44521", @@ -12829,14 +12902,15 @@ "issue:44933", "issue:44964", "issue:45127", + "issue:45245", "issue:45276", "issue:45357", "issue:45362", "issue:45375", "issue:45381", "issue:45520", + "issue:45563", "issue:45584", - "issue:45600", "issue:45632", "issue:46170", "issue:46369", @@ -12848,248 +12922,247 @@ "issue:46897", "issue:47145", "issue:47312", - "issue:47321", - "issue:47448" + "issue:47321" ], "soft_pairs": [ - "issue:43065|issue:44610", + "issue:39692|issue:47321", + "issue:43550|issue:44190", + "issue:45563|issue:45584", "issue:44617|issue:46369", - "issue:45381|issue:46655", + "issue:43827|issue:47321", + "issue:43065|issue:44610", "issue:44479|issue:44734", + "issue:45381|issue:46655", + "issue:41720|issue:45245", "issue:43526|issue:46697", - "issue:45632|issue:46617", "issue:43643|issue:45375", - "issue:43335|issue:44190", "issue:39692|issue:44190", + "issue:43335|issue:44190", "issue:43295|issue:45276", "issue:44779|issue:46489", + "issue:45632|issue:46617", "issue:44933|issue:45362", - "issue:43540|issue:44734", "issue:43854|issue:44964", + "issue:43540|issue:44734", "issue:45276|issue:45584", - "issue:43673|issue:46421", - "issue:44242|issue:46897", "issue:43994|issue:47321", "issue:45127|issue:45357", + "issue:44242|issue:46897", "issue:47145|issue:47321", - "issue:45520|issue:46170", + "issue:43673|issue:46421", "issue:43643|issue:44496", - "issue:44521|issue:45381", + "issue:45520|issue:46170", "issue:43317|issue:47312", - "issue:44112|issue:45600", - "issue:44279|issue:47448", - "issue:43611|issue:47321", - "issue:44016|issue:47448", - "issue:44112|issue:44964" + "issue:44521|issue:45381" ], - "prepared_review_unit_hash": "cf5b30e7fcef22db5c50c5640dd830dd3cd9828235cf01b8e851f25876d48ccd", + "prepared_review_unit_hash": "1499fdf516506527dea65384a14507b88cee1f1c35002f460100491d65d7ae2e", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36352, - "estimated_input_tokens": 9088, - "estimated_eval_tokens": 18432 + "serialized_chars": 35823, + "estimated_input_tokens": 8956, + "estimated_eval_tokens": 18168 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous set of unrelated model, processor, cache, and CI issues. The only plausible duplicate pair is the DeepSeek tokenizer regression (44779/46489); the rest differ by model, code path, or symptom.", - "confidence": 0.63, - "canonical_issue_reason": "issue:44779 is the clearest and broadest statement of the likely duplicate DeepSeek tokenizer regression, while 46489 reads as a narrower follow-up/gap in the same fix.", + "summary": "The cluster is largely heterogeneous: most items are unrelated regressions across tokenizers, multimodal processors, config loading, caching, and serving. I found one plausible duplicate pair in the DeepSeek tokenizer regression area; the rest are not the same underlying bug.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue is clearly canonical; the items span multiple unrelated code paths, so there is no strong duplicate anchor for the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44779 best represents the shared DeepSeek tokenizer regression because it describes the core v5-vs-v4 incorrect-output bug without the narrower PR-gap framing.", + "best_issue_reason": "No issue is a reliable global representative because the cluster lacks one shared root cause.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43065", - "right": "issue:44610", + "left": "issue:39692", + "right": "issue:47321", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and AMD quark class updates are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "A torch.compile/SDPA model bug is unrelated to a local dataset loading script failure." + }, + { + "left": "issue:45563", + "right": "issue:45584", "accept": false, - "reason": "Different SAM/OmDet topics and different failures: dummy Conv2d in a decoder vs processor/model input-size mismatch." + "reason": "Paged generate warning handling and Whisper empty-transcription failure are different generation bugs." }, { "left": "issue:44617", "right": "issue:46369", "accept": false, - "reason": "Both are memory-related, but they concern different SAM3 variants and different symptoms (OOM vs memory overflow), so not the same bug." + "reason": "Both are memory-related in Sam3, but they describe different model variants and failure modes, not the same bug." }, { - "left": "issue:45381", - "right": "issue:46655", + "left": "issue:43827", + "right": "issue:47321", + "accept": false, + "reason": "Docs referencing pipeline() after v5 removals is unrelated to AMD quark class updates." + }, + { + "left": "issue:43065", + "right": "issue:44610", "accept": false, - "reason": "Both involve video processing, but they are different models and likely different preprocessing bugs." + "reason": "Sam3PixelDecoder's dummy Conv2d and OmDet-Turbo image-size mismatch are different model/processor issues." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "One is a multimodal video-input regression, the other is a serve-time KV-cache tensor indexing crash; different code paths." + "reason": "A video-input regression in Qwen VL models is unrelated to a serving KV-cache tensor indexing crash." }, { - "left": "issue:43526", - "right": "issue:46697", + "left": "issue:45381", + "right": "issue:46655", "accept": false, - "reason": "Unrelated image-processing bugs: BEiT reduce_labels vs Swin2SR over-padding." + "reason": "Wrong vision_position_ids in qwen2.5-vl and SmolVLM video padding are different multimodal processor bugs." }, { - "left": "issue:45632", - "right": "issue:46617", + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA asserts and the 2^24 category limit are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:46697", "accept": false, - "reason": "Both touch trust_remote_code/cache loading, but one is a leaf-dir collision and the other is a FileNotFoundError from cache loading; not clearly the same underlying defect." + "reason": "BEiT label reduction and Swin2SR over-padding affect different processors and code paths." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Both are config-field issues, but one is a generic AutoConfig/trust_remote_code problem and the other is a specific Qwen vision config field dropped by @strict." + "reason": "Generic AutoConfig field loss under trust_remote_code and Qwen3.5MoeVisionConfig dropping one field are similar in theme but not the same concrete bug." }, { - "left": "issue:43335", + "left": "issue:39692", "right": "issue:44190", "accept": false, - "reason": "Completely different areas: SwitchTransformers config creation vs local dataset loading in an example script." + "reason": "SigLIP2 example errors are unrelated to loading a local dataset in the image-classification script." }, { - "left": "issue:39692", + "left": "issue:43335", "right": "issue:44190", "accept": false, - "reason": "SigLIP2 doc errors are unrelated to dataset loading in the image-classification example." + "reason": "SwitchTransformers sparse-layer creation and dataset loading in a training script are unrelated." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Different regressions: processor.tokenizer breakage vs resize_token_embeddings not propagating to Gemma4 embeddings." + "reason": "Processor.tokenizer regression and resize_token_embeddings behavior are different API issues." }, { "left": "issue:44779", "right": "issue:46489", "accept": true, - "reason": "Same underlying DeepSeek tokenizer regression in v5+; 46489 looks like a narrower follow-up to the incorrect-output bug described in 44779." + "reason": "Both report the same DeepSeek tokenizer regression in v5+ producing incorrect outputs; 46489 reads like a specific follow-up to the broader 44779 report." }, { - "left": "issue:44933", - "right": "issue:45362", + "left": "issue:45632", + "right": "issue:46617", "accept": false, - "reason": "Nonexistent image_utils import and Qwen chat crash are unrelated." + "reason": "Both involve trust_remote_code caching, but one is a leaf-directory collision and the other is a cache FileNotFoundError; not enough to treat as the same bug." }, { - "left": "issue:43540", - "right": "issue:44734", + "left": "issue:44933", + "right": "issue:45362", "accept": false, - "reason": "Different problems: Qwen3OmniMoe video-input validation vs transformers serve KV-cache indexing." + "reason": "A missing import from image_utils is unrelated to a Qwen3.5 chat crash." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both are model-loading failures, but for different models and different root causes; too broad to merge." + "reason": "These are separate model-loading failures for different models with no shared root cause shown." }, { - "left": "issue:45276", - "right": "issue:45584", - "accept": false, - "reason": "Gemma4 embedding resize bug and Whisper empty-transcription generation failure are unrelated." - }, - { - "left": "issue:43673", - "right": "issue:46421", + "left": "issue:43540", + "right": "issue:44734", "accept": false, - "reason": "Both mention chunked prefill/cache, but one is missing GenerationMixin cache in v5 and the other is lazy-init causing recompilation; not the same concrete bug." + "reason": "Qwen3OmniMoe video-input validation and serving KV-cache indexing are different code paths." }, { - "left": "issue:44242", - "right": "issue:46897", + "left": "issue:45276", + "right": "issue:45584", "accept": false, - "reason": "Different training-loss bugs in different models; only a vague pattern overlap." + "reason": "Embedding resize propagation and Whisper empty-transcription generation are unrelated bugs." }, { "left": "issue:43994", "right": "issue:47321", "accept": false, - "reason": "SigLIP2 wrong outputs and AMD quark class update are unrelated." + "reason": "SigLIP2 nonsensical outputs and AMD quark class updates are unrelated." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "LoRA/tied-embedding collapse and Qwen3.5 visual encoder save_pretrained keys are different subsystems." + "reason": "LoRA merge with extended vocabulary and saving Qwen3.5 visual encoder keys are different model-management issues." + }, + { + "left": "issue:44242", + "right": "issue:46897", + "accept": false, + "reason": "MoE load-balancing loss and Florence2 double-shift training loss are unrelated losses." }, { "left": "issue:47145", "right": "issue:47321", "accept": false, - "reason": "FP8 checkpoint weight_scale corruption is unrelated to AMD quark class updates." + "reason": "FP8 weight_scale loading and AMD quark class updates are unrelated." }, { - "left": "issue:45520", - "right": "issue:46170", + "left": "issue:43673", + "right": "issue:46421", "accept": false, - "reason": "Different platform-specific import failures: flash_attn on Python 3.13 vs torch.distributed dependency on Windows/ROCm." + "reason": "Chunked-prefill cache missing vs. chunked-prefill recompilation are related in theme but not clearly the same concrete defect." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "Both are config-related, but one is missing fields from trust_remote_code AutoConfig and the other is an unrecognized model_type in config.json." + "reason": "Missing fields from trust_remote_code config loading and an unrecognized model_type error are different config-loading failures." }, { - "left": "issue:44521", - "right": "issue:45381", + "left": "issue:45520", + "right": "issue:46170", "accept": false, - "reason": "Assistant-mask generation for multimodal templates is different from a Qwen video position-id regression." + "reason": "Python 3.13 flash_attn import handling and Windows ROCm torch.distributed dependency issues are unrelated." }, { "left": "issue:43317", "right": "issue:47312", "accept": false, - "reason": "Dequantized-model offload loading and kernel-version incompatibility are unrelated." - }, - { - "left": "issue:44112", - "right": "issue:45600", - "accept": false, - "reason": "Both are CI failures, but caused by different code and different regressions." - }, - { - "left": "issue:44279", - "right": "issue:47448", - "accept": false, - "reason": "Generic dependency issue vs Gemma4 architecture recognition problem; no shared bug." - }, - { - "left": "issue:43611", - "right": "issue:47321", - "accept": false, - "reason": "Base-model-prefix loading regression and AMD quark class updates are unrelated." - }, - { - "left": "issue:44016", - "right": "issue:47448", - "accept": false, - "reason": "Notebook syntax error and Gemma4 arch recognition are unrelated." + "reason": "Dequantized model offload loading and kernel version compatibility are separate problems." }, { - "left": "issue:44112", - "right": "issue:44964", + "left": "issue:44521", + "right": "issue:45381", "accept": false, - "reason": "GraniteSpeech CI flake and Phi-4 multimodal model loading failure are unrelated." + "reason": "All-zero assistant masks and wrong vision_position_ids are different multimodal template/positioning bugs." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall and conservative on the soft edges. The only accepted pair, issue:44779 with issue:46489, is well supported by the titles as the same DeepSeek tokenizer regression plus a narrower follow-up gap in the earlier fix. The other rejections stay within the evidence available from the packet and avoid overclaiming. " + "feedback": "The assessment is grounded in the packet and stays appropriately conservative. The cluster does appear heterogeneous from the titles alone, and the one accepted duplicate pair (44779/46489) is plausibly the same DeepSeek tokenizer regression. The other soft-edge rejections are cautious and well-supported by the visible issue descriptions." }, "evaluator_used": true, "retried": false, @@ -13117,18 +13190,18 @@ "issue:35141", "issue:36246", "issue:38617", - "issue:39692", "issue:41628", - "issue:41950", "issue:42175", "issue:42796", "issue:43295", "issue:43335", "issue:43404", "issue:43531", + "issue:43611", "issue:43673", "issue:43976", "issue:44016", + "issue:44112", "issue:44246", "issue:44279", "issue:44462", @@ -13151,233 +13224,246 @@ "issue:45803", "issue:45907", "issue:46234", - "issue:46424", "issue:46506", "issue:46752", "issue:46858", - "issue:46932", "issue:47217", + "issue:47321", "issue:47448" ], "soft_pairs": [ - "issue:43673|issue:46858", + "issue:44279|issue:47448", + "issue:43611|issue:47321", + "issue:44112|issue:45600", + "issue:44016|issue:47448", "issue:42175|issue:44279", + "issue:44112|issue:44964", + "issue:43673|issue:46858", "issue:44246|issue:45003", "issue:44462|issue:45907", "issue:27957|issue:35141", - "issue:44964|issue:45600", - "issue:32090|issue:35141", "issue:42796|issue:44521", + "issue:32090|issue:35141", + "issue:44964|issue:45600", "issue:44737|issue:45803", "issue:43404|issue:45216", "issue:43531|issue:45600", "issue:43335|issue:45593", + "issue:25251|issue:29925", "issue:44496|issue:47217", "issue:45290|issue:46752", - "issue:45632|issue:46234", "issue:44496|issue:45042", - "issue:44938|issue:45542", - "issue:36246|issue:38617", - "issue:36246|issue:41628", + "issue:45632|issue:46234", "issue:43976|issue:47448", - "issue:25251|issue:29925", + "issue:36246|issue:41628", + "issue:36246|issue:38617", "issue:43295|issue:45335", "issue:45083|issue:46506", - "issue:41950|issue:44016", "issue:44496|issue:45759", - "issue:43673|issue:46424", - "issue:39692|issue:46932" + "issue:44938|issue:45542" ], - "prepared_review_unit_hash": "f9f473886e618a29f3b35c8a0dbaed8e347a682322362655d38b895bd05c6e52", + "prepared_review_unit_hash": "811de344b87636eddef3ea5c32840cffebbedac44539673a336b9b432758c456", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 35933, - "estimated_input_tokens": 8984, - "estimated_eval_tokens": 18224 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35899, + "estimated_input_tokens": 8975, + "estimated_eval_tokens": 18206 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is heterogeneous: it mixes unrelated issues across cache handling, tokenizers, multimodal chat templates, import errors, model config bugs, and docs/CI breakages. None of the proposed soft pairs are strong duplicates of the same underlying bug/change.", + "summary": "This cluster is highly heterogeneous: the issues span unrelated model-loading, tokenizer, chat-template, packaging, CI, and training bugs. None of the soft pairs look like the same underlying defect, so no duplicates should be merged here.", "confidence": 0.94, - "canonical_issue_reason": "No single canonical issue stands out because the items are not one duplicate set; they span many unrelated subsystems and root causes.", + "canonical_issue_reason": "No single canonical issue; the cluster is not a true duplicate set and mixes many unrelated failure modes.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue: there isn\u2019t a coherent cluster around one defect, so picking a canonical issue would be arbitrary.", + "best_issue_reason": "No issue cleanly represents the cluster, since the items cover different subsystems and distinct root causes.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43673", - "right": "issue:46858", + "left": "issue:44279", + "right": "issue:47448", + "accept": false, + "reason": "Generic dependency/install complaint vs a specific Gemma4 architecture recognition bug." + }, + { + "left": "issue:43611", + "right": "issue:47321", + "accept": false, + "reason": "Base-model-prefix loading regression is unrelated to AMD quark class refresh." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "CI failure in GraniteSpeech device override is unrelated to stale auto_mappings entries." + }, + { + "left": "issue:44016", + "right": "issue:47448", "accept": false, - "reason": "Both are cache-related, but one is a missing cache in chunked prefill and the other is a device-tensor seq-length bug in static cache arithmetic; different failure modes." + "reason": "Notebook syntax error is documentation-related, not a Gemma4 model-loading issue." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Both mention dependencies, but one is about TensorFlow not being installed with a torch extra and the other is a generic transformer dependency issue; not the same bug." + "reason": "TensorFlow missing from the torch extra is a concrete packaging bug, not a vague dependency issue." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "GraniteSpeech CI flake and Phi-4 multimodal loading failure are different problems." + }, + { + "left": "issue:43673", + "right": "issue:46858", + "accept": false, + "reason": "Missing GenerationMixin cache in v5.0.0 is distinct from device-tensor slice arithmetic in prefill." }, { "left": "issue:44246", "right": "issue:45003", "accept": false, - "reason": "Import slowness and unsafe sys.modules access are distinct problems; one is performance, the other is correctness/safety." + "reason": "Slow import reports a performance issue; unsafe sys.modules access is a correctness bug." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "Tokenizer repository selection and list_repo_files revision forwarding are unrelated API issues." + "reason": "Tokenizer.json being ignored is unrelated to forwarding revision in list_repo_files." }, { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta is unrelated to embedding reinitialization after resize_token_embeddings." + "reason": "Flash Attention 2 support for XLM-R is unrelated to resize-token-embeddings reinitializing output weights." }, { - "left": "issue:44964", - "right": "issue:45600", + "left": "issue:42796", + "right": "issue:44521", "accept": false, - "reason": "One is a model-loading failure for Phi-4 multimodal; the other is stale auto_mappings references breaking CI. Different code paths." + "reason": "BPE marker handling in a tokenizer is different from assistant_mask generation for multimodal chat templates." }, { "left": "issue:32090", "right": "issue:35141", "accept": false, - "reason": "Trainer broadcast TypeError and post_init embedding reinitialization are unrelated bugs." + "reason": "Trainer broadcast NoneType error is unrelated to post-init embedding reinitialization." }, { - "left": "issue:42796", - "right": "issue:44521", + "left": "issue:44964", + "right": "issue:45600", "accept": false, - "reason": "Tokenizer BPE marker handling and apply_chat_template assistant_mask behavior are different tokenizer/chat-template issues." + "reason": "Phi-4 multimodal load failure and stale auto_mappings references are different load paths and causes." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "CPU-only positional encoding allocation and bare-except swallowing errors in Fuyu conversion are unrelated." + "reason": "CPU arange in XLNet is a performance/device bug, not a bare-except exception-handling issue." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "lm_head weight tying in Mistral3 and Qwen3.5 save_pretrained checkpoint correctness are different model-specific bugs." + "reason": "Untied lm_head weights in Mistral3 is distinct from Qwen3.5 checkpoint save corruption." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and removed Sam3LiteText config references are unrelated." + "reason": "Qwen3-MoE sliding_window behavior is unrelated to removed Sam3LiteText config references." }, { "left": "issue:43335", "right": "issue:45593", "accept": false, - "reason": "SwitchTransformers sparse-layer construction and D-FINE auxiliary-loss handling are separate model behaviors." + "reason": "Sparse-layer creation in SwitchTransformers is unrelated to D-FINE auxiliary loss behavior." + }, + { + "left": "issue:25251", + "right": "issue:29925", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to SigLIP fast-tokenizer enforcement." }, { "left": "issue:44496", "right": "issue:47217", "accept": false, - "reason": "Unrecognized model/config model_type and multimodal image-placeholder failures are distinct loading vs input-format issues." + "reason": "Unrecognized model type error is unrelated to VL image-placeholder handling in user text." }, { "left": "issue:45290", "right": "issue:46752", "accept": false, - "reason": "Both hit apply_chat_template, but one is about tool-call assistant messages and the other about an empty list edge case; not the same defect." - }, - { - "left": "issue:45632", - "right": "issue:46234", - "accept": false, - "reason": "Cache-path collisions for trust_remote_code and a missing availability helper are unrelated." + "reason": "Both hit apply_chat_template, but one is tool-call/no-content handling and the other is empty-input IndexError." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Model-type recognition and PIL image-processor torchvision dependency are different areas." + "reason": "Model-type resolution failure is unrelated to PIL image processors wrongly requiring torchvision." }, { - "left": "issue:44938", - "right": "issue:45542", + "left": "issue:45632", + "right": "issue:46234", "accept": false, - "reason": "Python 3.14 import failure and TensorBoard/TensorFlow backend misconfiguration are unrelated environment issues." + "reason": "Local cache path collisions are unrelated to a missing import_utils availability helper." }, { - "left": "issue:36246", - "right": "issue:38617", + "left": "issue:43976", + "right": "issue:47448", "accept": false, - "reason": "Missing Qwen2_5_VLImageProcessor export and missing layer_type_validation export are separate import errors." + "reason": "Python version compatibility breakage is unrelated to Gemma4 architecture registration." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they involve different symbols and different modules." - }, - { - "left": "issue:43976", - "right": "issue:47448", - "accept": false, - "reason": "Python version compatibility and Gemma4 architecture recognition are unrelated." + "reason": "Missing Qwen2_5_VLImageProcessor export is a different import failure than AutoImageProcessor absence." }, { - "left": "issue:25251", - "right": "issue:29925", + "left": "issue:36246", + "right": "issue:38617", "accept": false, - "reason": "Pipeline top_k nesting behavior and SigLIP use_fast enforcement do not share the same underlying bug." + "reason": "A missing processor export and a missing configuration_utils symbol are unrelated import errors." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Processor/tokenizer regression and t5gemma decoder.embed_tokens resize mismatch are different model integration bugs." + "reason": "Processor/tokenizer regression is unrelated to t5gemma resize_token_embeddings not affecting decoder embeddings." }, { "left": "issue:45083", "right": "issue:46506", "accept": false, - "reason": "qwen3_omni_moe feature-length helper semantics and CPU/CUDA output divergence in PPDocLayoutV3 are different issues." - }, - { - "left": "issue:41950", - "right": "issue:44016", - "accept": false, - "reason": "Video-classification pipeline processor selection and a notebook syntax error are unrelated." + "reason": "Qwen3_omni_moe feature-length helper behavior is unrelated to PPDocLayoutV3 CPU/CUDA divergence." }, - { - "left": "issue:44496", - "right": "issue:45759", - "accept": false, - "reason": "Missing model_type in config and AutoModelForCausalLM.from_config not unwrapping text_config are different config-resolution bugs." - }, - { - "left": "issue:43673", - "right": "issue:46424", + { + "left": "issue:44496", + "right": "issue:45759", "accept": false, - "reason": "Both involve cache, but one is a missing cache during chunked prefill and the other is max_cache_length being ignored outside quantized cache." + "reason": "Model-type recognition failure is not the same as unwrapping text_config for composite multimodal configs." }, { - "left": "issue:39692", - "right": "issue:46932", + "left": "issue:44938", + "right": "issue:45542", "accept": false, - "reason": "Documentation example failures for SigLIP2 and Gemma4 quantized-embedder dtype casting are unrelated." + "reason": "Python 3.14 import failure is unrelated to tensorboard installing without TensorFlow." } ] }, @@ -13409,7 +13495,6 @@ "issue:42907", "issue:42994", "issue:43295", - "issue:43299", "issue:43381", "issue:43441", "issue:43526", @@ -13432,7 +13517,6 @@ "issue:44933", "issue:44987", "issue:45020", - "issue:45216", "issue:45335", "issue:45405", "issue:45446", @@ -13440,249 +13524,260 @@ "issue:45588", "issue:45593", "issue:45812", + "issue:45823", "issue:46082", "issue:46121", "issue:46424", "issue:46506", "issue:46515", "issue:46752", + "issue:46932", "issue:47217", "issue:47332", "issue:47333" ], "soft_pairs": [ + "issue:39692|issue:46932", + "issue:41950|issue:44016", "issue:43976|issue:44987", + "issue:44279|issue:46515", "issue:44683|issue:45446", "issue:42994|issue:47333", - "issue:43673|issue:44336", "issue:29925|issue:33290", - "issue:44279|issue:46515", + "issue:43673|issue:46424", + "issue:43673|issue:44336", "issue:43381|issue:45593", - "issue:44295|issue:46506", "issue:43526|issue:43901", - "issue:44327|issue:46752", + "issue:44295|issue:46506", + "issue:44016|issue:44220", "issue:46082|issue:46121", + "issue:44327|issue:46752", "issue:43650|issue:44220", - "issue:44016|issue:44220", "issue:42994|issue:47332", "issue:44016|issue:44062", "issue:43441|issue:45588", "issue:44792|issue:46424", - "issue:42907|issue:43441", "issue:41950|issue:45020", + "issue:42907|issue:43441", "issue:45335|issue:45584", "issue:43824|issue:44933", "issue:44279|issue:45405", + "issue:39692|issue:43901", "issue:43901|issue:44496", "issue:43295|issue:44779", - "issue:39692|issue:43901", "issue:44779|issue:45812", - "issue:43901|issue:47217", "issue:42796|issue:43650", - "issue:43299|issue:45216" + "issue:43901|issue:47217", + "issue:41950|issue:45823" ], - "prepared_review_unit_hash": "bf3199687d466594432b26142fabd786bfbe83d2126f61ec151dd74cdee50a55", + "prepared_review_unit_hash": "1220ee94181e8b725af5ad9d3b9b038c67dbfd652cdedc7aacc4fab21814fd15", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36032, - "estimated_input_tokens": 9008, - "estimated_eval_tokens": 18272 + "soft_pair_count": 32, + "serialized_chars": 36916, + "estimated_input_tokens": 9229, + "estimated_eval_tokens": 18714 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is not a duplicate cluster: it mixes unrelated Transformers issues across tokenizers, model loading/saving, attention backends, pipelines, configs, docs, and environment compatibility. None of the soft pairs look like the same concrete bug/change.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue fits the whole set; the reports are heterogeneous. If a representative is needed, issue:45020 is the broadest version-regression tracker, but it does not subsume the rest.", + "summary": "This cluster is heterogeneous: the titles point to many unrelated bugs, regressions, and doc issues across different models/subsystems. I rejected all soft edges; none look like the same underlying change or bug.", + "confidence": 0.94, + "canonical_issue_reason": "#45020 is the broadest, most general regression report in the set and is better suited as a representative issue than the narrower model- or feature-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the broadest umbrella-like report in the set (recent version regressions with remote_code models), so it is the closest thing to a representative issue.", + "best_issue_reason": "#45020 has the strongest cluster-representative value: it is a clearly described, high-signal regression with broader scope than the other issues.", "best_pr_reason": null, "soft_edge_verdicts": [ + { + "left": "issue:39692", + "right": "issue:46932", + "accept": false, + "reason": "Different models and different failures: SigLIP2 doc example errors vs Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Unrelated bugs: a video-classification processor lookup issue vs a notebook syntax error." + }, { "left": "issue:43976", "right": "issue:44987", "accept": false, - "reason": "Different failures: Python version compatibility vs a specific model-loading regression." + "reason": "Both involve version regressions, but one is a Python support issue and the other is a model-loading failure for a specific repo; not the same bug." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "A dependency issue with transformers is not the same as a torch version incompatibility." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both mention flex_attention, but one is a torch>=2.9 compile failure and the other is a bad version check for AuxRequest import." + "reason": "Same subsystem (flex_attention) but different concrete problems: compiled flex_attention on torch>=2.9 vs a bad version check for AuxRequest import." }, { "left": "issue:42994", "right": "issue:47333", "accept": false, - "reason": "Both are save/offload problems, but they involve different model families and different save paths." + "reason": "Both are save failures, but they involve different model families and different save contexts (dequantized vs disk-offloaded)." }, { - "left": "issue:43673", - "right": "issue:44336", + "left": "issue:29925", + "right": "issue:33290", "accept": false, - "reason": "Unrelated components: GenerationMixin chunked_prefill cache vs ANSI output in loading_report." + "reason": "Tokenizer fast-mode enforcement is unrelated to an Adafactor/DeepSpeed OOM." }, { - "left": "issue:29925", - "right": "issue:33290", + "left": "issue:43673", + "right": "issue:46424", "accept": false, - "reason": "Tokenizer fast-mode enforcement is unrelated to a deepspeed/Adafactor OOM." + "reason": "Both mention cache, but one is a missing GenerationMixin cache in chunked prefill and the other is cache-length behavior for quantized cache only." }, { - "left": "issue:44279", - "right": "issue:46515", + "left": "issue:43673", + "right": "issue:44336", "accept": false, - "reason": "Generic dependency complaint vs a specific torch 2.4.1 incompatibility; not the same bug." + "reason": "Cache-missing regression and ANSI output formatting are unrelated." }, { "left": "issue:43381", "right": "issue:45593", "accept": false, - "reason": "Gradient-checkpointing eval-mode behavior is unrelated to D-FINE auxiliary-loss handling." + "reason": "Gradient checkpointing in eval mode is unrelated to D-FINE auxiliary-loss behavior." }, { - "left": "issue:44295", - "right": "issue:46506", + "left": "issue:43526", + "right": "issue:43901", "accept": false, - "reason": "A buffer-reading error for position_ids is unrelated to CPU/CUDA divergence in PPDocLayoutV3." + "reason": "Different components and bug types: BeitImageProcessorFast label reduction vs documentation drift in TextClassificationPipeline." }, { - "left": "issue:43526", - "right": "issue:43901", + "left": "issue:44295", + "right": "issue:46506", "accept": false, - "reason": "Different areas: BeitImageProcessorFast label reduction vs documentation wording for text classification." + "reason": "A position_ids buffer access error is unrelated to CPU/CUDA output differences in PPDocLayoutV3." }, { - "left": "issue:44327", - "right": "issue:46752", + "left": "issue:44016", + "right": "issue:44220", "accept": false, - "reason": "Both are edge-case index errors, but they occur in different APIs and code paths." + "reason": "A notebook syntax error is unrelated to the fbank feature extraction bug." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "One is LlamaConfig head_dim validation; the other is rope-parameter conversion of a JSON-loaded list." + "reason": "Both are config-validation issues, but they concern different fields and different code paths." }, { - "left": "issue:43650", - "right": "issue:44220", + "left": "issue:44327", + "right": "issue:46752", "accept": false, - "reason": "Placeholder title is unrelated to the audio feature extraction issue." + "reason": "Both are edge-case crashes, but in different APIs and with different failure modes." }, { - "left": "issue:44016", + "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "Notebook syntax error and fbank feature extraction are unrelated." + "reason": "The placeholder issue 'ADD THE DATA' is not the same as the fbank feature extraction failure." }, { "left": "issue:42994", "right": "issue:47332", "accept": false, - "reason": "Both are save/offloading issues, but the reported contexts and model families differ enough that they do not look like one concrete bug." + "reason": "Same general saving area, but distinct models and distinct offloading/save scenarios." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Notebook syntax error is unrelated to the AddedToken special-argument TypeError." + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both mention flash_attention, but one is a Ministral model failure and the other is an s_aux=None crash in a different code path." + "reason": "Both touch flash attention, but one is a model-specific v5 RC failure and the other is an integration AttributeError on sink-less models." }, { "left": "issue:44792", "right": "issue:46424", "accept": false, - "reason": "Janus test failure is unrelated to cache max_length behavior for Quantized Cache." + "reason": "A janus test failure is unrelated to cache-length semantics." }, { - "left": "issue:42907", - "right": "issue:43441", + "left": "issue:41950", + "right": "issue:45020", "accept": false, - "reason": "Saving dequantized models is unrelated to a flash-attention model-loading failure." + "reason": "Video-classification processor lookup and remote_code loading regressions are different bugs." }, { - "left": "issue:41950", - "right": "issue:45020", + "left": "issue:42907", + "right": "issue:43441", "accept": false, - "reason": "Video-classification pipeline processor lookup is a specific pipeline bug, not the remote_code regression described in 45020." + "reason": "Saving dequantized Ministrals/Devstrals is unrelated to Ministral-3 FlashAttention failures." }, { "left": "issue:45335", "right": "issue:45584", "accept": false, - "reason": "resize_token_embeddings on t5gemma is unrelated to Whisper generation after align_special_tokens." + "reason": "Tokenizer embedding resize behavior and Whisper empty transcription generation are different issues." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols in different modules." + "reason": "Both are import-related, but they concern different missing symbols in different modules." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "Generic dependency issue vs a specific unreleased PEFT version bump; not the same change." + "reason": "A transformers dependency issue is unrelated to an unreleased PEFT minimum version bump." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "A SigLIP2 docs/model mismatch is unrelated to a pipeline documentation note about return_all_scores." }, { "left": "issue:43901", "right": "issue:44496", "accept": false, - "reason": "Documentation drift for text-classification is unrelated to an unrecognized model/config loading failure." + "reason": "Documentation drift in a pipeline API is unrelated to an unrecognized-model config error." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different model families and different failure modes." - }, - { - "left": "issue:39692", - "right": "issue:43901", - "accept": false, - "reason": "SigLIP2 example errors are not the same issue as stale text-classification docs." - }, - { - "left": "issue:44779", - "right": "issue:45812", - "accept": false, - "reason": "Different tokenizer regressions for different model families; not one concrete fix." + "reason": "Both are regression reports around tokenizers, but they describe different models and different breakages." }, { "left": "issue:43901", "right": "issue:47217", "accept": false, - "reason": "Docs mentioning return_all_scores is unrelated to VL placeholder handling in user text." - }, - { - "left": "issue:42796", - "right": "issue:43650", - "accept": false, - "reason": "Ministral tokenizer BPE-marker handling is unrelated to the placeholder issue." + "reason": "A pipeline docs issue is unrelated to VL models failing on image placeholders in user text." }, { - "left": "issue:43299", - "right": "issue:45216", + "left": "issue:41950", + "right": "issue:45823", "accept": false, - "reason": "One is a model-loading regression, the other is a save_pretrained checkpoint regression; different code paths." + "reason": "Video-classification processor lookup and Gemma4 device-map mismatch are different code paths and failures." } ] }, @@ -13706,7 +13801,6 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ - "issue:41950", "issue:43299", "issue:43381", "issue:43479", @@ -13715,6 +13809,7 @@ "issue:43650", "issue:43653", "issue:43854", + "issue:43873", "issue:43931", "issue:43976", "issue:44016", @@ -13739,7 +13834,7 @@ "issue:45542", "issue:45588", "issue:45593", - "issue:45823", + "issue:45800", "issue:45859", "issue:45907", "issue:46082", @@ -13756,224 +13851,238 @@ "issue:47405" ], "soft_pairs": [ - "issue:41950|issue:45823", + "issue:43299|issue:45216", "issue:43479|issue:45375", + "issue:44466|issue:45276", "issue:46123|issue:46369", "issue:46614|issue:47085", - "issue:44466|issue:45276", "issue:43577|issue:45216", "issue:43653|issue:45593", - "issue:43976|issue:45405", "issue:47145|issue:47405", + "issue:43976|issue:45405", "issue:43381|issue:47000", "issue:44462|issue:45081", "issue:46082|issue:46399", - "issue:44075|issue:44368", "issue:43299|issue:45357", - "issue:46082|issue:46512", + "issue:43299|issue:43854", "issue:43644|issue:47145", + "issue:44075|issue:44368", + "issue:46082|issue:46512", "issue:44242|issue:45200", - "issue:43299|issue:43854", "issue:44016|issue:44291", - "issue:43931|issue:47145", "issue:44683|issue:45542", + "issue:43931|issue:47145", "issue:43650|issue:46209", "issue:45588|issue:47000", - "issue:44351|issue:45278", - "issue:45859|issue:46851", "issue:45081|issue:45907", - "issue:45103|issue:45372" + "issue:45859|issue:46851", + "issue:45103|issue:45372", + "issue:43873|issue:47145", + "issue:44351|issue:45278", + "issue:45542|issue:45800" ], - "prepared_review_unit_hash": "fe4f882ccbf9e926e0b28f8d858acbb9cae6e3ab26864409e22dcba9e2f3a826", + "prepared_review_unit_hash": "45b4e926bacd89f73536fe4757321e39d1fbe4e5ab61f3a3cd4c232b9cde45de", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 36195, - "estimated_input_tokens": 9049, - "estimated_eval_tokens": 18354 + "soft_pair_count": 29, + "serialized_chars": 36773, + "estimated_input_tokens": 9194, + "estimated_eval_tokens": 18644 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a loose topical cluster of Transformers regressions across unrelated code paths (tokenizers, multimodal configs, loading/saving, losses, offload, and docs). None of the soft pairs look like true duplicates, so I would not merge any of them.", - "confidence": 0.28, - "canonical_issue_reason": "issue:45278 is the broadest regression report in the set and best captures the cluster\u2019s general theme of version-upgrade breakage, even though the individual items are not duplicates.", + "summary": "This cluster is heterogeneous: most items are distinct model/config/load/save/tokenizer bugs rather than duplicates. No soft pair looks like the same underlying change or failure mode, so none should be merged together.", + "confidence": 0.93, + "canonical_issue_reason": "No true canonical duplicate stands out; if one must be chosen, issue 43299 is the broadest early regression report and the closest thing to a hub, but it is still distinct from the others.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the most useful umbrella representative because it is a user-facing upgrade regression and can contextualize the surrounding import/load failures better than the narrower model-specific bugs.", + "best_issue_reason": "Issue 43299 is the most central candidate only in the sense that it is an early, user-facing regression report with several superficial similarity links; however it does not actually subsume the other issues.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:41950", - "right": "issue:45823", + "left": "issue:43299", + "right": "issue:45216", "accept": false, - "reason": "Different bugs: image-processor lookup in video classification vs a Gemma4 device-mapping mismatch during forward." + "reason": "Both mention Qwen regressions, but one is a loading failure and the other is incorrect save_pretrained serialization; different code paths and symptoms." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Both are config-related, but they affect different models and different fields/default-initialization paths." + "reason": "Different models and different config bugs: Phi4Multimodal default init vs Qwen3_5MoeVisionConfig field dropping." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "One is tied-weight serialization varying by device, the other is Gemma 4 embedding resize propagation; not the same defect." }, { "left": "issue:46123", "right": "issue:46369", "accept": false, - "reason": "MaskGenerationPipeline batch-tracking bug is unrelated to sam3-litetext memory overflow." + "reason": "MaskGenerationPipeline batch-finalization bug vs sam3-litetext memory overflow; unrelated failures." }, { "left": "issue:46614", "right": "issue:47085", "accept": false, - "reason": "Stale phonemizer backend state and tokenizer thread-safety are different failure modes in different subsystems." - }, - { - "left": "issue:44466", - "right": "issue:45276", - "accept": false, - "reason": "One is serialization of tied lm_head weights; the other is resize_token_embeddings not propagating into Gemma4-specific embeddings." + "reason": "Stale phonemizer backend state is unrelated to Rust tokenizer thread safety." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "BLIP2 dtype casting and Qwen3.5 save_pretrained checkpoint corruption are separate model-loading/save regressions." + "reason": "Blip2 dtype propagation bug and Qwen3.5 checkpoint save regression are unrelated." }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "Tokenizer special-token registration and D-FINE auxiliary-loss gating are unrelated bugs." + "reason": "Tokenizer special-token registration bug vs D-FINE auxiliary-loss behavior; no shared underlying bug." }, { - "left": "issue:43976", - "right": "issue:45405", + "left": "issue:47145", + "right": "issue:47405", "accept": false, - "reason": "Python version support breakage is not the same as an unreleased PEFT minimum-version bump." + "reason": "Both are silent load/corruption issues, but one is FP8 weight_scale application and the other is wrong Auto class/model_type conversion skip; different mechanics." }, { - "left": "issue:47145", - "right": "issue:47405", + "left": "issue:43976", + "right": "issue:45405", "accept": false, - "reason": "Missing FP8 weight_scale application is a different bug from silent checkpoint conversion skipping under the wrong Auto class." + "reason": "Python version support issue vs unreleased MIN_PEFT_VERSION bump; unrelated." }, { "left": "issue:43381", "right": "issue:47000", "accept": false, - "reason": "Gradient checkpointing in eval mode and RTDetrLoss infeasible cost matrix are unrelated code paths." + "reason": "Gradient checkpointing eval-mode restriction is unrelated to RTDetrLoss infeasible cost matrix errors." }, { "left": "issue:44462", "right": "issue:45081", "accept": false, - "reason": "Tokenizer.json selection and Mistral regex patching are distinct tokenizer-loading issues." + "reason": "AutoTokenizer ignoring tokenizer.json is a separate loading bug from Mistral regex patch crashing on missing backend_tokenizer." }, { "left": "issue:46082", "right": "issue:46399", "accept": false, - "reason": "LlamaConfig head_dim validation and Qwen3.5 rotary-kernel registration are different model/config bugs." - }, - { - "left": "issue:44075", - "right": "issue:44368", - "accept": false, - "reason": "SGD optimizer args being ignored is unrelated to a tied-word-embeddings warning in Qwen3.5 fine-tuning." + "reason": "LlamaConfig head_dim validation and Qwen3.5 rotary-kernel registration failure are different issues." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Qwen3VL MoE loading failure and Qwen3.5 save_pretrained visual-key regression are different model/version issues." + "reason": "Both are Qwen3.5/Qwen3VL regressions, but one is model loading and the other is save_pretrained visual key serialization; not the same bug." }, { - "left": "issue:46082", - "right": "issue:46512", + "left": "issue:43299", + "right": "issue:43854", "accept": false, - "reason": "Head-dim validation and stale model.config.dtype after from_config are separate configuration bugs." + "reason": "Different Qwen variants and different failure modes: model loading mismatch vs unit-test load failure." }, { "left": "issue:43644", "right": "issue:47145", "accept": false, - "reason": "Non-persistent buffers being junk-filled is not the same as FP8 compressed-tensors missing weight_scale on load." + "reason": "Non-persistent buffer junk in Transformers 5.0 is unrelated to FP8 compressed-tensors weight_scale loading." }, { - "left": "issue:44242", - "right": "issue:45200", + "left": "issue:44075", + "right": "issue:44368", "accept": false, - "reason": "Missing load-balancing loss and Gemma 4 mm_token_type_ids defaults are different training-behavior bugs." + "reason": "Optimizer SGD args ignored and tie_word_embeddings warning are unrelated; no shared code-path." }, { - "left": "issue:43299", - "right": "issue:43854", + "left": "issue:46082", + "right": "issue:46512", "accept": false, - "reason": "These are different model-loading failures for different models and root causes." + "reason": "Both involve model config parameters, but one is Llama head_dim validation and the other is from_config dtype staleness." }, { - "left": "issue:44016", - "right": "issue:44291", + "left": "issue:44242", + "right": "issue:45200", "accept": false, - "reason": "Notebook syntax error and init_empty_weights TypeError are unrelated failures." + "reason": "Load-balancing loss gating and Gemma 4 mm_token_type_ids defaulting are unrelated." }, { - "left": "issue:43931", - "right": "issue:47145", + "left": "issue:44016", + "right": "issue:44291", "accept": false, - "reason": "Qwen3-VL weight shape mismatch is not the same as FP8 checkpoint loading missing weight_scale." + "reason": "Notebook syntax error and init_empty_weights unexpected argument are unrelated." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Compiled flex_attention on torch>=2.9 and the tensorboard-without-TF backend error are unrelated." + "reason": "Torch 2.9 flex_attention compilation failure is unrelated to tensorboard/TensorFlow backend installation behavior." + }, + { + "left": "issue:43931", + "right": "issue:47145", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch and FP8 weight_scale omission are different loading failures." }, { "left": "issue:43650", "right": "issue:46209", "accept": false, - "reason": "Placeholder issue title versus an FP8 inference-on-B200 bug; no shared underlying defect." + "reason": "Placeholder/non-technical issue vs FP8 inference failure on B200; no relation." }, { "left": "issue:45588", "right": "issue:47000", "accept": false, - "reason": "Flash-attention crash on s_aux=None and RTDetrLoss infeasible cost matrix are different components and failures." + "reason": "Flash-attention auxiliary state crash and RTDetrLoss infeasible cost matrix are different subsystems and bugs." }, { - "left": "issue:44351", - "right": "issue:45278", + "left": "issue:45081", + "right": "issue:45907", "accept": false, - "reason": "Missing HybridCache import and broad import errors after upgrade are related only at a very high level, not the same bug." + "reason": "Mistral regex tokenizer crash vs list_repo_files revision forwarding; unrelated." }, { "left": "issue:45859", "right": "issue:46851", "accept": false, - "reason": "CPU offload incompatibility in rotary embedding and variable-length position_ids handling are different Qwen3.5 issues." + "reason": "Both are Qwen3.5 rotary/position-related, but one is CPU offload compatibility and the other is variable-length position_ids handling; not the same concrete bug." }, { - "left": "issue:45081", - "right": "issue:45907", + "left": "issue:45103", + "right": "issue:45372", "accept": false, - "reason": "Tokenizer regex patching crash and list_repo_files revision forwarding are unrelated." + "reason": "auto_docstring annotation handling and mistral_common import failure are unrelated." }, { - "left": "issue:45103", - "right": "issue:45372", + "left": "issue:43873", + "right": "issue:47145", + "accept": false, + "reason": "Quantization/offloading behavior and FP8 frozen checkpoint weight_scale application are distinct loading paths." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "A single missing HybridCache import is not the same as broad import errors after upgrade." + }, + { + "left": "issue:45542", + "right": "issue:45800", "accept": false, - "reason": "Auto-docstring future-annotations crash and Gemma 4 processor import failure stem from different parts of the stack." + "reason": "TensorBoard without TensorFlow backend error and torch/transformers incompatibility are unrelated." } ] }, @@ -14006,14 +14115,15 @@ "issue:43479", "issue:43525", "issue:43575", + "issue:43643", "issue:43673", + "issue:43723", "issue:43761", "issue:43784", "issue:43824", - "issue:43873", - "issue:43881", "issue:43901", "issue:43976", + "issue:44162", "issue:44186", "issue:44188", "issue:44246", @@ -14035,7 +14145,6 @@ "issue:45600", "issue:45656", "issue:45712", - "issue:45800", "issue:45834", "issue:46399", "issue:46846", @@ -14047,251 +14156,238 @@ "issue:47427" ], "soft_pairs": [ - "issue:43873|issue:47145", "issue:43441|issue:45834", "issue:42994|issue:43317", "issue:44336|issue:46851", - "issue:45542|issue:45800", - "issue:43784|issue:45542", "issue:44671|issue:44779", - "issue:44246|issue:47217", + "issue:43784|issue:45542", "issue:44186|issue:47000", + "issue:44246|issue:47217", "issue:42175|issue:43976", - "issue:44623|issue:44749", "issue:45325|issue:46851", "issue:43673|issue:43761", - "issue:43479|issue:43525", - "issue:33290|issue:45656", - "issue:43525|issue:43881", - "issue:44623|issue:45397", + "issue:44623|issue:44749", "issue:46399|issue:46846", - "issue:43901|issue:44246", - "issue:44261|issue:44263", + "issue:43525|issue:43643", + "issue:43479|issue:43525", "issue:44589|issue:45362", + "issue:43901|issue:44246", + "issue:44623|issue:45397", "issue:42796|issue:43441", + "issue:33290|issue:45656", "issue:43575|issue:47145", + "issue:44261|issue:44263", "issue:44188|issue:44336", "issue:45600|issue:45712", "issue:45588|issue:47427", "issue:45362|issue:47333", "issue:43824|issue:43976", + "issue:44246|issue:44496", "issue:44336|issue:44373", - "issue:44246|issue:44496" + "issue:43723|issue:44162" ], - "prepared_review_unit_hash": "117c7e5f1bba778370f85e6c9f5c1c74f108778c3e1151e6952df45ff988f0f0", + "prepared_review_unit_hash": "8eaf943b6ea439728105dd94115c2e7a5c665a4830249ecbb9887019da0e80a7", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36480, - "estimated_input_tokens": 9120, - "estimated_eval_tokens": 18496 + "soft_pair_count": 29, + "serialized_chars": 36258, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "A heterogeneous set of unrelated Transformers issues clustered by broad embedding similarity, not by a shared underlying bug. The soft links mostly reflect the same broad areas (offloading, quantization, model loading, tokenizer/model config, docs), but not mergeable duplicates. No PRs are present in the cluster.", - "confidence": 0.93, - "canonical_issue_reason": "No single true duplicate hub stands out; issue 43873 is the best representative because it is the only open item and the most discussed/referenced, with a broad offloading+quantization scope closest to the cluster\u2019s general theme.", + "summary": "This cluster is highly heterogeneous: most items are unrelated model-, tokenizer-, packaging-, or runtime-specific bugs that only share broad Transformers/v5 wording. I do not see a safe duplicate merge candidate among the soft edges.", + "confidence": 0.21, + "canonical_issue_reason": "If a representative issue is needed, issue 44246 is the broadest/highest-signal runtime problem in the set, but the overall cluster is not a coherent duplicate family.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43873 is the most suitable anchor for triage because it is active, has the most discussion/inbound references, and is broad enough to represent the cluster even though the cluster is not a real duplicate set.", + "best_issue_reason": "Issue 44246 is the most representative general-runtime complaint among otherwise unrelated reports, though it should not be treated as a true cluster duplicate anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43873", - "right": "issue:47145", - "accept": false, - "reason": "Both involve offloading/quantization, but one is about generic offloading behavior and the other is a specific FP8 compressed-tensors weight_scale omission; different code paths and fixes." - }, { "left": "issue:43441", "right": "issue:45834", "accept": false, - "reason": "Same model family, but one is a FlashAttention regression and the other is a long-ocr index error; unrelated failure modes." + "reason": "Different bugs: Ministral tokenizer BPE-marker handling vs Kosmos2.5 OCR long-input index error." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Both touch quantized/dequantized model handling, but saving failure and gpu+cpu offload loading failure are distinct bugs." + "reason": "Saving a quantized model and loading a dequantized model with offload are different code paths and failures." }, { "left": "issue:44336", "right": "issue:46851", "accept": false, - "reason": "One is ANSI codes in loading_report, the other is Qwen3.5 variable-length position_ids handling; no shared underlying bug." - }, - { - "left": "issue:45542", - "right": "issue:45800", - "accept": false, - "reason": "Packaging/backend dependency issue versus a torch/transformers version incompatibility; too broad and unrelated." - }, - { - "left": "issue:43784", - "right": "issue:45542", - "accept": false, - "reason": "Import NameError in sentence-transformers integration is unrelated to TensorBoard/TensorFlow backend installation behavior." + "reason": "ANSI-code leakage in loading_report is unrelated to Qwen3.5 variable-length position_ids handling." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "Different models and different symptom classes: masked LM correctness vs tokenizer output regression." + "reason": "CamemBERT masked-LM regression and Deepseek tokenizer corruption are distinct model/tokenizer problems." }, { - "left": "issue:44246", - "right": "issue:47217", + "left": "issue:43784", + "right": "issue:45542", "accept": false, - "reason": "Slow import and VL image-placeholder handling are unrelated issues." + "reason": "Both are import/backend errors, but they affect different packages and stem from different missing symbols/backends." }, { "left": "issue:44186", "right": "issue:47000", "accept": false, - "reason": "Different model components and different errors: tokenizer batching crash vs RTDetr loss infeasible cost matrix." + "reason": "Tokenizer crash on NER/padding is not the same underlying bug as RTDetrLoss infeasible cost matrix." }, { - "left": "issue:42175", - "right": "issue:43976", + "left": "issue:44246", + "right": "issue:47217", "accept": false, - "reason": "Backend installation confusion versus Python version compatibility; not the same bug." + "reason": "Import-time slowness and VL image-placeholder failure are unrelated behaviors." }, { - "left": "issue:44623", - "right": "issue:44749", + "left": "issue:42175", + "right": "issue:43976", "accept": false, - "reason": "Processor save-pretrained missing files and Chinese performance regression are unrelated." + "reason": "Package extra dependency inclusion differs from Python-version compatibility; not the same defect." }, { "left": "issue:45325", "right": "issue:46851", "accept": false, - "reason": "Both are Qwen position-id issues, but one is a still-image rope scaling bug in Qwen2.5-VL and the other is variable-length input handling in Qwen3.5; different concrete code paths." + "reason": "Qwen2.5-VL rope-index scaling and Qwen3.5 variable-length position_ids are separate model bugs." }, { "left": "issue:43673", "right": "issue:43761", "accept": false, - "reason": "Cache missing in generation/chunked prefill is distinct from CLIPVisionModel hidden_states regression." + "reason": "Missing generation cache in chunked_prefill is unrelated to CLIPVision hidden_states regression." }, { - "left": "issue:43479", - "right": "issue:43525", + "left": "issue:44623", + "right": "issue:44749", "accept": false, - "reason": "Multimodal config default initialization and missing pad_token_id are separate config bugs." + "reason": "Missing save_pretrained files and slower filtering after upgrade are different symptoms and code paths." }, { - "left": "issue:33290", - "right": "issue:45656", + "left": "issue:46399", + "right": "issue:46846", "accept": false, - "reason": "Both mention Deepspeed/optimizers, but OOM with Adafactor and double optimizer.step calls are different problems." + "reason": "A kernel-registration crash from a missing decorator is not the same as TP-plan OOM/shape mismatch." }, { "left": "issue:43525", - "right": "issue:43881", + "right": "issue:43643", "accept": false, - "reason": "Missing pad_token_id in one config does not match a GLM-4V loading failure." + "reason": "A missing pad_token_id attribute in Llama4Config is different from trust_remote_code dropping fields in AutoConfig." }, { - "left": "issue:44623", - "right": "issue:45397", + "left": "issue:43479", + "right": "issue:43525", "accept": false, - "reason": "Processor save files issue and gemma-4 zero3 from_pretrained failure are not the same underlying bug." + "reason": "Multimodal default-config initialization and missing config attributes are unrelated configuration bugs." }, { - "left": "issue:46399", - "right": "issue:46846", + "left": "issue:44589", + "right": "issue:45362", "accept": false, - "reason": "Kernel registration/decorator error and missing base_model_tp_plan entries causing TP OOM are distinct issues." + "reason": "Float8 storage lookup failure and Qwen3.5 chat crash do not share a concrete code path." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Docs mismatch for text classification pipeline is unrelated to intermittent slow imports." + "reason": "A documentation mismatch is not the same as import performance degradation." }, { - "left": "issue:44261", - "right": "issue:44263", + "left": "issue:44623", + "right": "issue:45397", "accept": false, - "reason": "Precision drift from missing rms_norm_eps and torch.split return-value handling are different code-path bugs." + "reason": "Processor save output omissions and gemma-4 zero3 from_pretrained failures are distinct issues." }, { - "left": "issue:44589", - "right": "issue:45362", + "left": "issue:42796", + "right": "issue:43441", "accept": false, - "reason": "Missing Float8 storage object and Qwen3.5 chat crash are unrelated failures." + "reason": "Tokenizer BPE-marker parsing and FlashAttention failure are different parts of the stack." }, { - "left": "issue:42796", - "right": "issue:43441", + "left": "issue:33290", + "right": "issue:45656", "accept": false, - "reason": "Tokenizer BPE marker handling and FlashAttention crash in Ministral-3 are different bugs." + "reason": "AdaFactor OOM in deepspeed and double optimizer-step behavior are not the same underlying bug." }, { "left": "issue:43575", "right": "issue:47145", "accept": false, - "reason": "TP load OOM and FP8 compressed-tensors save/load corruption are different memory/model-loading problems." + "reason": "Tensor-parallel load OOM and FP8 checkpoint weight_scale omission are different loading bugs." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "MLA q_a_layernorm epsilon handling is unrelated to torch.split return-value behavior." }, { "left": "issue:44188", "right": "issue:44336", "accept": false, - "reason": "Torch.compile attention-kernel divergence and loading_report ANSI output are unrelated." + "reason": "torch.compile-dependent attention-kernel divergence has nothing to do with ANSI output in loading_report." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Removed auto-mapping configs breaking CI and leftover dummy classes in dummy_pt_objects are separate cleanup/test failures." + "reason": "Stale auto_mappings entries and leftover dummy_pt_objects cleanup are separate maintenance failures." }, { "left": "issue:45588", "right": "issue:47427", "accept": false, - "reason": "FlashAttention sink-less-model crash and redundant initialization skipping on non-CUDA hardware are different code paths." + "reason": "FlashAttention sink-less-model crash and redundant init on non-CUDA hardware are different initialization paths." }, { "left": "issue:45362", "right": "issue:47333", "accept": false, - "reason": "Qwen3.5 chat crash and disk-offloaded save failure are unrelated." + "reason": "Qwen3.5 chat crashes and disk-offloaded checkpoint saving are unrelated failures." }, { "left": "issue:43824", "right": "issue:43976", "accept": false, - "reason": "Missing exported model class and Python version support issue are unrelated." + "reason": "An import error for a missing class and a Python support/package issue are not duplicates." }, { - "left": "issue:44336", - "right": "issue:44373", + "left": "issue:44246", + "right": "issue:44496", "accept": false, - "reason": "ANSI color leakage and docstring wording are not the same bug." + "reason": "Import slowness is unrelated to unrecognized-model/config.json validation failure." }, { - "left": "issue:44246", - "right": "issue:44496", + "left": "issue:44336", + "right": "issue:44373", "accept": false, - "reason": "Slow import and unrecognized model config are different failures." + "reason": "ANSI-code generation and a wrong docstring are not the same issue." }, { - "left": "issue:44261", - "right": "issue:44496", + "left": "issue:43673", + "right": "issue:44162", "accept": false, - "reason": "Layernorm precision bug and model config recognition error are unrelated." + "reason": "Generation cache regression and ESM2 breakage are entirely different model/runtime bugs." } ] }, @@ -14320,13 +14416,11 @@ "issue:42915", "issue:43257", "issue:43299", + "issue:43352", "issue:43526", "issue:43540", "issue:43550", "issue:43596", - "issue:43723", - "issue:44162", - "issue:44242", "issue:44246", "issue:44295", "issue:44297", @@ -14339,13 +14433,14 @@ "issue:45083", "issue:45103", "issue:45137", + "issue:45250", "issue:45292", "issue:45375", "issue:45397", "issue:45561", "issue:45563", "issue:45588", - "issue:45593", + "issue:45701", "issue:45759", "issue:46291", "issue:46399", @@ -14365,12 +14460,11 @@ "issue:47312" ], "soft_pairs": [ - "issue:43723|issue:44162", + "issue:42491|issue:43299", "issue:46291|issue:46399", "issue:46399|issue:46851", - "issue:42491|issue:43299", - "issue:46655|issue:46726", "issue:45072|issue:46829", + "issue:46655|issue:46726", "issue:45083|issue:46851", "issue:44297|issue:45397", "issue:47176|issue:47177", @@ -14380,230 +14474,231 @@ "issue:45003|issue:46821", "issue:45103|issue:46399", "issue:44246|issue:44871", + "issue:43550|issue:46506", "issue:45561|issue:45563", "issue:44336|issue:46424", "issue:44295|issue:45397", - "issue:43550|issue:46506", "issue:43540|issue:46752", "issue:42915|issue:47312", - "issue:43540|issue:46655", "issue:36246|issue:43596", + "issue:43540|issue:46655", "issue:46614|issue:46665", - "issue:46726|issue:47217", "issue:44479|issue:46655", - "issue:43257|issue:45137", - "issue:44779|issue:45292", + "issue:46726|issue:47217", "issue:45375|issue:45759", - "issue:44242|issue:45593" + "issue:43257|issue:45137", + "issue:43352|issue:45250", + "issue:45292|issue:45701", + "issue:44779|issue:45292" ], - "prepared_review_unit_hash": "a106c6780339f97280600767ce1229e8cef17d6aae77ffa6b7e4c46e6273099c", + "prepared_review_unit_hash": "41e7f001a0ce52873c9f033f53014f5ea2f666be4fad3788bf2b2168d78dee4f", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 30, - "serialized_chars": 36819, - "estimated_input_tokens": 9205, - "estimated_eval_tokens": 18666 + "serialized_chars": 36198, + "estimated_input_tokens": 9050, + "estimated_eval_tokens": 18356 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly heterogeneous transformer bug reports with only superficial lexical similarity. The soft-similarity links do not indicate true duplicates, so none of the candidate pairs should be merged. The only loosely related pocket is Qwen multimodal/video-loading issues, but even there the failure modes and code paths differ.", - "confidence": 0.91, - "canonical_issue_reason": "issue:44479 is the broadest concrete regression in the only loosely related subset (Qwen VL/video handling), so it is the best anchor if one must choose a representative issue.", + "summary": "These items are mostly separate bugs across different model families and subsystems; none of the soft pairs looks like a safe duplicate merge on the same concrete code path. The only strong commonality is broad Qwen/VL regression chatter, but the reported failures are still materially different.", + "confidence": 0.74, + "canonical_issue_reason": "issue:44479 is the broadest concrete regression report here: it names a specific version breakage and multiple affected Qwen VL/MoE variants, making it the best anchor for nearby vision-regression reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 has the clearest user-facing reproduction and the widest model coverage among the related Qwen multimodal reports.", + "best_issue_reason": "issue:44479 is the best representative issue in this set because it is self-contained, version-specific, and spans multiple related model variants; the others are either narrower, unrelated, or clearly different bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43723", - "right": "issue:44162", + "left": "issue:42491", + "right": "issue:43299", "accept": false, - "reason": "Tokenizer loading failure vs ESM2 breakage; different subsystems and symptoms." + "reason": "Both are Qwen3 MoE loading regressions, but one is a LoRA compatibility problem and the other is a base-model load failure; not the same concrete bug." }, { "left": "issue:46291", "right": "issue:46399", "accept": false, - "reason": "Pipelines import/version check issue vs Qwen3.5 rotary-registration bug; unrelated code paths." + "reason": "One is a kernels import/version requirement failure, the other is a Qwen3.5 kernel registration error; same area, different failure mode and fix." }, { "left": "issue:46399", "right": "issue:46851", "accept": false, - "reason": "Missing decorator on rotary embedding registration vs variable-length position_ids handling; different bugs." + "reason": "Different problems in Qwen3.5: kernels registration vs variable-length position_ids handling; not mergeable as one bug." }, { - "left": "issue:42491", - "right": "issue:43299", + "left": "issue:45072", + "right": "issue:46829", "accept": false, - "reason": "Both involve Qwen3 MoE loading, but one is a hf4.x/hf5.x LoRA compatibility issue and the other is accelerate+deepspeed weight conversion." + "reason": "Unrelated model/path issues: bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel versus VideoPrism output formatting." }, { "left": "issue:46655", "right": "issue:46726", "accept": false, - "reason": "Video padding calculation error vs missing-vision-backend processor resolution; different failure layers." - }, - { - "left": "issue:45072", - "right": "issue:46829", - "accept": false, - "reason": "Bfloat16 dtype mismatch in SwitchTransformers/TimmWrapper vs VideoPrism output field mix-up; unrelated." + "reason": "Both involve vision processing, but one is a video padding bug and the other is missing-backend AutoProcessor error handling." }, { "left": "issue:45083", "right": "issue:46851", "accept": false, - "reason": "Helper output-length computation in qwen3_omni_moe vs position_ids handling in Qwen3.5; not the same bug." + "reason": "Both touch Qwen input-shape logic, but one is feature-extraction length math and the other is position_ids with variable-length inputs." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer save-pretrained metadata mismatch vs gemma-4 zero3 from_pretrained failure; different areas." + "reason": "Tokenizer save metadata mismatch is unrelated to Gemma-4 ZeRO3 from_pretrained behavior." }, { "left": "issue:47176", "right": "issue:47177", "accept": false, - "reason": "Both are security issues, but command-substitution bypass and tar/zip slip are distinct vulnerabilities." + "reason": "Both are security issues, but command-substitution allowlist bypass and tar/zip slip arbitrary write are distinct vulnerabilities with different code paths." }, { "left": "issue:45003", "right": "issue:47217", "accept": false, - "reason": "Unsafe sys.modules access in modeling_utils vs VL image-placeholder parsing; unrelated." + "reason": "sys.modules access in modeling_utils is unrelated to VL placeholder handling in user text." }, { "left": "issue:43526", "right": "issue:45003", "accept": false, - "reason": "BEiT image processor label reduction bug vs modeling_utils module lookup issue; no shared code path." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to unsafe sys.modules access." }, { "left": "issue:45588", "right": "issue:46851", "accept": false, - "reason": "Flash-attention crash on s_aux=None vs variable-length position_ids bug; unrelated." + "reason": "Flash-attention sinkless-model crash and Qwen3.5 variable-length position_ids are different defects." }, { "left": "issue:45003", "right": "issue:46821", "accept": false, - "reason": "sys.modules access bug vs TimesFM window_size AttributeError; unrelated." + "reason": "Generic modeling_utils sys.modules access does not match TimesFM window_size AttributeError." }, { "left": "issue:45103", "right": "issue:46399", "accept": false, - "reason": "auto_docstring annotations crash vs Qwen3.5 rotary decorator issue; different functionality." + "reason": "auto_docstring future-annotations crash is unrelated to Qwen3.5 kernel registration." }, { "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Import latency report vs Gemma eos_token_id inconsistency; not the same defect." + "reason": "Import-time performance regression and Gemma eos_token_id inconsistency are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba is unrelated to CPU/CUDA divergence in PPDocLayoutV3." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "xdist file race vs stale warning in paged generate; different behavior and scope." + "reason": "pytest-xdist file race and stale generate() warning are different test/behavior issues." }, { "left": "issue:44336", "right": "issue:46424", "accept": false, - "reason": "ANSI output formatting vs cache max length handling; unrelated." + "reason": "ANSI coloring in loading_report and cache max length handling are unrelated." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "position_ids buffer access error vs gemma-4 zero3 loading; different models and code paths." - }, - { - "left": "issue:43550", - "right": "issue:46506", - "accept": false, - "reason": "torch.compile SDPA failure in Bamba vs CPU/CUDA divergence in PPDocLayoutV3; unrelated." + "reason": "position_ids buffer access and Gemma-4 ZeRO3 from_pretrained are separate problems." }, { "left": "issue:43540", "right": "issue:46752", "accept": false, - "reason": "Qwen3OmniMoe video ValueError vs empty-list chat template crash; unrelated." + "reason": "Qwen3OmniMoe video-input ValueError and empty chat list IndexError are unrelated." }, { "left": "issue:42915", "right": "issue:47312", "accept": false, - "reason": "Qwen3Moe FineGrainedFP8Config failure vs kernels 0.16 incompatibility; different causes." + "reason": "FineGrainedFP8Config failure and kernels 0.16 incompatibility are different versioning/runtime issues." }, { - "left": "issue:43540", - "right": "issue:46655", + "left": "issue:36246", + "right": "issue:43596", "accept": false, - "reason": "Both are video-related, but one is a Qwen3OmniMoe input validation failure and the other is a SmolVLM padding bug." + "reason": "Qwen2_5_VLImageProcessor import error and BertModel ZeRO3 index error are unrelated." }, { - "left": "issue:36246", - "right": "issue:43596", + "left": "issue:43540", + "right": "issue:46655", "accept": false, - "reason": "Qwen2_5_VL import error vs DeepSpeed ZeRO3 BertModel IndexError; unrelated." + "reason": "Both are video-related, but they affect different models and different processing bugs." }, { "left": "issue:46614", "right": "issue:46665", "accept": false, - "reason": "Stale phonemizer backend state vs from_pretrained inconsistency inside classes; different mechanisms." + "reason": "Stale phonemizer backend state is unrelated to inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44479", + "right": "issue:46655", + "accept": false, + "reason": "Both are video-input regressions, but one is a broad Qwen version regression and the other is a SmolVLM video padding bug." }, { "left": "issue:46726", "right": "issue:47217", "accept": false, - "reason": "Missing vision backend error vs image-placeholder-in-user-text failure; both VL-adjacent but distinct." + "reason": "Missing vision backend handling is unrelated to VL placeholder placement in user text." }, { - "left": "issue:44479", - "right": "issue:46655", + "left": "issue:45375", + "right": "issue:45759", "accept": false, - "reason": "Both mention video input, but 44479 is a broad Qwen v5.3.0 regression while 46655 is a specific SmolVLM padding-calculation bug." + "reason": "Missing deepstack_visual_indexes and text_config unwrapping are separate Qwen multimodal config issues." }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MoE weight conversion issue vs ZeRO3 deque underflow; different subsystems." + "reason": "Both involve DeepSpeed/ZeRO3, but one is Qwen3 MoE weight conversion and the other is a deque underflow during initialization." }, { - "left": "issue:44779", - "right": "issue:45292", + "left": "issue:43352", + "right": "issue:45250", "accept": false, - "reason": "Deepseek tokenizer correctness vs resize_token_embeddings not updating output embeddings; unrelated." + "reason": "A model not yet supporting Flash Attention 2.0 is not the same as the generic Flash Attention 2.0 issue." }, { - "left": "issue:45375", - "right": "issue:45759", + "left": "issue:45292", + "right": "issue:45701", "accept": false, - "reason": "Missing config field in Qwen3.5MoeVisionConfig vs composite-config unwrapping in AutoModelForCausalLM.from_config; different bugs." + "reason": "Token embedding resizing and tokenizer-version drift are unrelated." }, { - "left": "issue:44242", - "right": "issue:45593", + "left": "issue:44779", + "right": "issue:45292", "accept": false, - "reason": "Load-balancing loss gating vs D-FINE auxiliary-loss omission; only loosely analogous, not the same change." + "reason": "Deepseek tokenizer regression and resize_token_embeddings/output_embeddings mismatch are different bugs." } ] }, @@ -14631,7 +14726,6 @@ "issue:42757", "issue:42886", "issue:43295", - "issue:43352", "issue:43475", "issue:43540", "issue:44038", @@ -14641,6 +14735,7 @@ "issue:44242", "issue:44295", "issue:44351", + "issue:44368", "issue:44451", "issue:44466", "issue:44479", @@ -14648,21 +14743,21 @@ "issue:44749", "issue:44857", "issue:44877", + "issue:44977", "issue:44987", "issue:44995", "issue:45003", "issue:45137", "issue:45200", - "issue:45250", "issue:45292", "issue:45335", "issue:45375", "issue:45459", "issue:45561", "issue:45588", + "issue:45593", "issue:45600", "issue:45685", - "issue:45701", "issue:46132", "issue:46170", "issue:46369", @@ -14677,256 +14772,253 @@ "issue:47455" ], "soft_pairs": [ - "issue:45292|issue:45701", - "issue:43352|issue:45250", - "issue:44995|issue:45200", + "issue:44242|issue:45593", "issue:43295|issue:45292", - "issue:36246|issue:44351", + "issue:44995|issue:45200", + "issue:45685|issue:46431", "issue:44060|issue:44077", - "issue:36246|issue:42757", "issue:46170|issue:47321", "issue:44162|issue:45200", "issue:44466|issue:45335", - "issue:42886|issue:46132", + "issue:36246|issue:44351", + "issue:36246|issue:42757", "issue:45137|issue:45600", + "issue:42886|issue:46132", "issue:44995|issue:45588", "issue:44451|issue:46752", - "issue:44295|issue:44749", - "issue:45685|issue:46431", - "issue:47217|issue:47429", "issue:44479|issue:46170", + "issue:44295|issue:44749", "issue:44877|issue:45375", + "issue:47217|issue:47429", "issue:43475|issue:46369", - "issue:45003|issue:45561", "issue:44877|issue:46170", + "issue:45003|issue:45561", "issue:44987|issue:45459", - "issue:47312|issue:47455", "issue:44623|issue:46726", - "issue:44038|issue:46506", + "issue:47312|issue:47455", "issue:43475|issue:43540", - "issue:44242|issue:44857" + "issue:44242|issue:44857", + "issue:44038|issue:46506", + "issue:44368|issue:44977" ], - "prepared_review_unit_hash": "1a9fb3b62e5eaaaeb034fdfea85ea3a13486e7d78bd24f1b4b1607fc3b2be839", + "prepared_review_unit_hash": "60d5d030ccbcb5ee435a69905eb8b571be9e2346b254003801cb16c629ce264f", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35848, - "estimated_input_tokens": 8962, - "estimated_eval_tokens": 18180 + "serialized_chars": 36000, + "estimated_input_tokens": 9000, + "estimated_eval_tokens": 18256 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue set. I found one plausible duplicate pair around `resize_token_embeddings` not updating tied embeddings, but the rest are distinct regressions/import errors/config bugs.", - "confidence": 0.83, - "canonical_issue_reason": "No single issue cleanly represents the whole cluster; if a representative is needed, 44479 is the broadest concrete regression report (v5.3.0 video-input failures across multiple Qwen VL variants).", + "summary": "This cluster is mostly heterogeneous: most items are unrelated model- or subsystem-specific regressions, not true duplicates. The only plausible duplicate pair is the kernels version-checking reports.", + "confidence": 0.63, + "canonical_issue_reason": "issue:47455 is the cleanest root-cause report: it points to a broken kernel version check, which can explain the compatibility failure raised in issue:47312.", "canonical_pr_reason": null, - "best_issue_reason": "44479 is the strongest anchor because it describes a widespread, version-specific regression with multiple affected model families and a clear reproducer surface.", + "best_issue_reason": "issue:47455 is the best representative issue because it describes the underlying version-check bug more directly than the symptom report.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:45292", - "right": "issue:45701", + "left": "issue:44242", + "right": "issue:45593", "accept": false, - "reason": "`resize_token_embeddings` vs tokenization/version-dependent tokenization behavior; different failure modes and code paths." + "reason": "Different models and loss paths; both mention missing losses, but they are not the same bug or code path." }, { - "left": "issue:43352", - "right": "issue:45250", + "left": "issue:43295", + "right": "issue:45292", "accept": false, - "reason": "One is a model-specific Flash Attention support error; the other is only a generic Flash Attention 2.0 topic, not the same bug." + "reason": "One is a processor/tokenizer regression, the other is embedding resizing not propagating; different failures." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "Stale cache crash in one MoE model vs missing `mm_token_type_ids` default in Gemma 4; unrelated issues." - }, - { - "left": "issue:43295", - "right": "issue:45292", - "accept": false, - "reason": "Processor/tokenizer regression with images vs embedding resize not updating output embeddings; different concrete bugs." + "reason": "Unrelated model bugs: stale cache on second forward vs missing mm_token_type_ids default." }, { - "left": "issue:36246", - "right": "issue:44351", + "left": "issue:45685", + "right": "issue:46431", "accept": false, - "reason": "Both are import errors, but for different missing symbols in different modules; not the same underlying breakage." + "reason": "MPS histogram/int error and missing chat template are unrelated processor/runtime issues." }, { "left": "issue:44060", "right": "issue:44077", "accept": false, - "reason": "Tied-weights warning bug vs `patchtsmixer` post-init policy; no shared code-path problem." - }, - { - "left": "issue:36246", - "right": "issue:42757", - "accept": false, - "reason": "Different missing imports from different packages/modules; too unrelated to be duplicates." + "reason": "Both are warning/config complaints, but they involve different models and different checks." }, { "left": "issue:46170", "right": "issue:47321", "accept": false, - "reason": "Windows/ROCm torch.distributed import issue vs AMD quark class update issue; different subsystems and symptoms." + "reason": "Different AMD-related problems: torch.distributed dependency vs quark class update." }, { "left": "issue:44162", "right": "issue:45200", "accept": false, - "reason": "ESM2 breakage vs Gemma 4 fine-tuning token-type default issue; unrelated model bugs." + "reason": "ESM2 breakage and Gemma 4 token-type defaults are unrelated model-specific regressions." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Both mention tied embeddings, but one is device-dependent serialization and the other is resize propagation to decoder embeddings; not the same concrete bug." + "reason": "Both involve embeddings, but one is serialization of tied weights while the other is resize_token_embeddings propagation; not the same concrete bug." }, { - "left": "issue:42886", - "right": "issue:46132", + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Different missing import errors for different symbols and packages." + }, + { + "left": "issue:36246", + "right": "issue:42757", "accept": false, - "reason": "Offline cache loading failure vs attention registration behavior change; unrelated." + "reason": "Separate ImportErrors from different modules; no shared root cause." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error vs stale auto-mapping reference in CI; unrelated." + "reason": "DeepSpeed deque failure and auto_mappings CI breakage are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and AttentionInterface.register behavior change are different issues." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "MoE stale indexer cache crash vs flash_attention `s_aux=None` attribute error; different code paths." + "reason": "Stale indexer cache vs flash-attention s_aux=None crash are unrelated." }, { "left": "issue:44451", "right": "issue:46752", "accept": false, - "reason": "Model loading failure vs empty chat-template input crash; unrelated." + "reason": "Model loading failure and empty-list chat template crash are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Video-input regression in VL models is unrelated to Windows/ROCm distributed import issues." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "Position buffer read error vs Chinese report of preprocessing slowdown; no shared bug signal." + "reason": "Position_ids buffer access error and performance slowdown after upgrade are different problems." }, { - "left": "issue:45685", - "right": "issue:46431", + "left": "issue:44877", + "right": "issue:45375", "accept": false, - "reason": "MPS histogram kernel limitation vs `apply_chat_template` missing template error; unrelated." + "reason": "Both mention strict config, but one blocks loading a config while the other silently drops a field; not clearly the same bug." }, { "left": "issue:47217", "right": "issue:47429", "accept": false, - "reason": "VL image-placeholder handling bug vs tokenizers dependency upper-bound issue; different layers entirely." + "reason": "VL image-placeholder failure and tokenizers version upper-bound conflict are unrelated." }, { - "left": "issue:44479", - "right": "issue:46170", + "left": "issue:43475", + "right": "issue:46369", "accept": false, - "reason": "Video-input regression in Qwen VL models vs Windows/ROCm import failure; unrelated." + "reason": "Same SAM3 family, but missing attribute and memory overflow are different failure modes." }, { "left": "issue:44877", - "right": "issue:45375", - "accept": false, - "reason": "Both involve strict config handling, but they affect different model configs and failure modes; not the same concrete bug." - }, - { - "left": "issue:43475", - "right": "issue:46369", + "right": "issue:46170", "accept": false, - "reason": "SAM3 video attribute error vs SAM3-litetext memory overflow; different symptoms and likely different fixes." + "reason": "Strict config loading and Windows/ROCm distributed import dependency are unrelated." }, { "left": "issue:45003", "right": "issue:45561", "accept": false, - "reason": "`sys.modules` access bug vs xdist race on captured file; unrelated testing/runtime issues." - }, - { - "left": "issue:44877", - "right": "issue:46170", - "accept": false, - "reason": "Strict config loading vs Windows/ROCm import dependency issue; unrelated." + "reason": "sys.modules access and xdist file race are different infrastructure bugs." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Model load failure vs tokenizer error masking when protobuf is absent; different problems." - }, - { - "left": "issue:47312", - "right": "issue:47455", - "accept": false, - "reason": "Kernels version compatibility vs incorrect version check; related theme but not the same bug." + "reason": "One is model loading on transformers>=5.1.0, the other is tokenizer error masking when protobuf is absent; different code paths." }, { "left": "issue:44623", "right": "issue:46726", "accept": false, - "reason": "Missing save_pretrained files vs AutoProcessor vision-backend fallback behavior; unrelated." + "reason": "Processor save-pretrained file omissions and AutoProcessor backend-missing errors are not the same bug." }, { - "left": "issue:44038", - "right": "issue:46506", - "accept": false, - "reason": "Qwen3-VL-Moe bug vs CPU/CUDA result divergence in PPDocLayoutV3; unrelated." + "left": "issue:47312", + "right": "issue:47455", + "accept": true, + "reason": "Both point to the same kernel package version-checking problem: one reports incompatibility with kernels 0.16, the other says the version check is wrong." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "Different model families and different failure modes (SAM3 video attr error vs Qwen3OmniMoe video processing ValueError)." + "reason": "Both are about SAM/Qwen model families, but one is a missing attribute and the other is a video-input ValueError." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "Load-balancing loss logic vs AMP/CUDA loss crash; no shared code-path problem." + "reason": "Load-balancing loss omission and LwDetr float16 AMP crash are unrelated training bugs." }, { - "left": "issue:45292", - "right": "issue:45335", - "accept": true, - "reason": "Same underlying `resize_token_embeddings` regression: both report the resize call failing to propagate to tied/output/decoder embeddings, with 45335 as a model-specific repro." + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3-VL-Moe version bug and PPDocLayoutV3 CPU/CUDA mismatch are different model issues." }, { - "left": "issue:44451", - "right": "issue:46752", + "left": "issue:44368", + "right": "issue:44977", "accept": false, - "reason": "Model loading failure vs empty-list chat-template crash; unrelated." + "reason": "Both mention Qwen3.5, but one is tied-weights warnings during fine-tuning and the other is flash-attention generation failure." }, { "left": "issue:45561", - "right": "issue:47217", + "right": "issue:47455", + "accept": false, + "reason": "Test utility file-race and kernel version-check logic are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45588", "accept": false, - "reason": "xdist race on captured info file vs VL image-placeholder parsing failure; unrelated." + "reason": "Video-input regression in VL models and flash_attention s_aux crash are different layers and failures." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it treats the cluster as mostly heterogeneous and only elevates the 47312/47455 pair, which is the only clearly related duplicate-like pair from the titles. The soft-edge verdicts are generally cautious and do not overstate evidence." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -14964,7 +15056,6 @@ "issue:44368", "issue:44483", "issue:44749", - "issue:44977", "issue:45103", "issue:45229", "issue:45397", @@ -14993,245 +15084,231 @@ "issue:47429" ], "soft_pairs": [ + "issue:46305|issue:46315", "issue:29870|issue:30333", - "issue:44368|issue:44977", - "issue:45397|issue:45656", "issue:43653|issue:44162", - "issue:46305|issue:46315", + "issue:45397|issue:45656", "issue:43723|issue:46710", + "issue:42886|issue:45229", "issue:43065|issue:43408", + "issue:45987|issue:46082", "issue:45412|issue:46123", - "issue:42886|issue:45229", - "issue:47059|issue:47429", "issue:43906|issue:46202", - "issue:45987|issue:46082", - "issue:43931|issue:44368", "issue:45103|issue:46858", - "issue:42199|issue:46851", - "issue:44361|issue:44749", + "issue:43931|issue:44368", + "issue:47059|issue:47429", + "issue:45593|issue:47068", "issue:43994|issue:46726", + "issue:44361|issue:44749", + "issue:42199|issue:46851", "issue:44368|issue:45563", + "issue:46752|issue:47020", "issue:46132|issue:47145", "issue:44188|issue:46202", - "issue:46752|issue:47020", - "issue:45593|issue:47068", - "issue:44483|issue:46202", - "issue:43976|issue:47429", + "issue:43784|issue:46399", "issue:43701|issue:44749", + "issue:43976|issue:47429", "issue:44188|issue:46421", - "issue:43784|issue:46399", - "issue:44749|issue:45656", - "issue:43065|issue:46123", - "issue:39692|issue:46726" + "issue:44483|issue:46202", + "issue:39692|issue:46726", + "issue:44749|issue:45656" ], - "prepared_review_unit_hash": "36df1dfce711fc5fd464a97bab0f2fad7ef2b9abc88c0d9db44a1418adf71fd8", + "prepared_review_unit_hash": "df7dcc4d705293ca8404f5a7eeaa2beafef8edf6c8e8350a1bc47ffabfd93fcf", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 37051, - "estimated_input_tokens": 9263, - "estimated_eval_tokens": 18782 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35903, + "estimated_input_tokens": 8976, + "estimated_eval_tokens": 18208 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: every soft pair appears to be a different bug/regression in a different code path, so none should be merged as duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No canonical issue: the items span unrelated subsystems and failure modes, with no single underlying bug tying them together.", + "summary": "These issues are mostly unrelated one-off bugs across tokenizers, generation, vision models, cache behavior, and training/inference tooling. None of the soft pairs look like the same underlying change or defect, so I would not merge any pair as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "No canonical duplicate stands out: the set is heterogeneous, and the soft pairs do not share a single concrete bug or code path.", "canonical_pr_reason": null, - "best_issue_reason": "No strong global representative exists; picking any one issue would be arbitrary because the cluster is not a duplicate group.", + "best_issue_reason": "issue:46305 is the strongest standalone report here: it has a precise off-by-one in `sliding_window_mask_function`, a clear symptom, and an actionable fix path.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:29870", - "right": "issue:30333", - "accept": false, - "reason": "Unrelated failures: BART generation hits an UnboundLocalError, while the other is an MLflow job-status reporting issue." - }, - { - "left": "issue:44368", - "right": "issue:44977", + "left": "issue:46305", + "right": "issue:46315", "accept": false, - "reason": "Both mention Qwen3.5, but one is a config warning about tied embeddings and the other is a flash-attention generation failure." + "reason": "MoonshineStreaming masking off-by-one vs tokenizer `convert_tokens_to_ids` slowdown; different subsystems and failures." }, { - "left": "issue:45397", - "right": "issue:45656", + "left": "issue:29870", + "right": "issue:30333", "accept": false, - "reason": "Different problems: Gemma-4 from_pretrained/ZeRO3 loading versus deepspeed double optimizer stepping." + "reason": "Bart `generate()` UnboundLocalError vs MLflow reporting after failed training; unrelated bugs." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "Tokenizer special-token registration bug versus an ESM2 model regression; no shared code-path." + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 model breakage; not the same code path." }, { - "left": "issue:46305", - "right": "issue:46315", + "left": "issue:45397", + "right": "issue:45656", "accept": false, - "reason": "One is a sliding-window mask off-by-one; the other is a tokenizer performance regression." + "reason": "Gemma-4 ZeRO-3 `from_pretrained` loading issue vs Deepspeed optimizer stepping twice; different problems." }, { "left": "issue:43723", "right": "issue:46710", "accept": false, - "reason": "Tokenizer-loading regression in v5 versus an accuracy regression for a specific model after upgrade." + "reason": "Tokenizer loading issue in v5 vs DeepSeek accuracy regression after upgrade; distinct symptoms and fixes." }, { - "left": "issue:43065", - "right": "issue:43408", + "left": "issue:42886", + "right": "issue:45229", "accept": false, - "reason": "Both are SAM3-related, but one is a dummy Conv2d implementation detail and the other is a model-type mismatch warning." + "reason": "Offline tokenizer cache loading failure vs Gemma4 multi-GPU CUDA OOM." }, { - "left": "issue:45412", - "right": "issue:46123", + "left": "issue:43065", + "right": "issue:43408", "accept": false, - "reason": "Memory not released on deletion is unrelated to the pipeline batch-finalization bug." + "reason": "Dummy Conv2d in Sam3PixelDecoder vs sam3_video/sam3_tracker warning; same family, different issue." }, { - "left": "issue:42886", - "right": "issue:45229", + "left": "issue:45987", + "right": "issue:46082", "accept": false, - "reason": "Offline tokenizer cache loading and Gemma multi-GPU CUDA OOM are different failures." + "reason": "Cache return-type contract bug vs LlamaConfig head_dim validation; unrelated." }, { - "left": "issue:47059", - "right": "issue:47429", + "left": "issue:45412", + "right": "issue:46123", "accept": false, - "reason": "CLI breakage from a hub API change is unrelated to the tokenizers version upper-bound packaging issue." + "reason": "RT-DETR memory not released on deletion vs MaskGenerationPipeline final-batch handling." }, { "left": "issue:43906", "right": "issue:46202", "accept": false, - "reason": "An isolated reproduction of an earlier issue is not the same bug as Gemma pipeline roundtrip failure on \u2581." + "reason": "Reproduction of another issue vs Gemma pipeline roundtrip failure on `\u2581`; no shared bug." }, { - "left": "issue:45987", - "right": "issue:46082", + "left": "issue:45103", + "right": "issue:46858", "accept": false, - "reason": "StaticCache return-type contract bug versus LlamaConfig validation of explicit head_dim; different paths." + "reason": "auto_docstring annotation crash vs static-cache seq-length tensor arithmetic; different code paths." }, { "left": "issue:43931", "right": "issue:44368", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch is unrelated to a Qwen3.5 embedding-tie warning." - }, - { - "left": "issue:45103", - "right": "issue:46858", - "accept": false, - "reason": "Auto-docstring annotation handling bug versus cache seq-length/device-tensor arithmetic bug." + "reason": "Qwen3-VL weight-shape mismatch vs tie_word_embeddings warning; not the same underlying defect." }, { - "left": "issue:42199", - "right": "issue:46851", + "left": "issue:47059", + "right": "issue:47429", "accept": false, - "reason": "DETR cardinality calculation versus Qwen3.5 variable-length position_ids; no common root cause." + "reason": "CLI completion regression with huggingface_hub vs tokenizers version upper-bound packaging issue." }, { - "left": "issue:44361", - "right": "issue:44749", + "left": "issue:45593", + "right": "issue:47068", "accept": false, - "reason": "MLLukeTokenizer AttributeError is unrelated to the v4.57.3\u2192v5.3.0 data-filtering slowdown." + "reason": "D-FINE auxiliary-loss behavior vs Deepspeed host-synchronization in loss computation." }, { "left": "issue:43994", "right": "issue:46726", "accept": false, - "reason": "SigLIP2 wrong outputs and AutoProcessor missing-backend error are distinct vision-model issues." + "reason": "SigLIP2 docs/example model-processor mismatch and quantization failure vs missing-vision-backend AutoProcessor error." }, { - "left": "issue:44368", - "right": "issue:45563", + "left": "issue:44361", + "right": "issue:44749", "accept": false, - "reason": "Qwen3.5 tie_word_embeddings warning is unrelated to the stale num_return_sequences warning in paged generate()." + "reason": "MLLukeTokenizer AttributeError on tasks vs filtering slowdown after Transformers upgrade." }, { - "left": "issue:46132", - "right": "issue:47145", + "left": "issue:42199", + "right": "issue:46851", "accept": false, - "reason": "AttentionInterface registration behavior change and FP8 weight_scale loading corruption are different bugs." + "reason": "DETR cardinality error for no-background-class models vs Qwen3.5 variable-length `position_ids` bug." }, { - "left": "issue:44188", - "right": "issue:46202", + "left": "issue:44368", + "right": "issue:45563", "accept": false, - "reason": "torch.compile attention-kernel divergence does not match the Gemma \u2581 roundtrip failure." + "reason": "Tie-word-embeddings warning during fine-tuning vs stale `num_return_sequences` warning." }, { "left": "issue:46752", "right": "issue:47020", "accept": false, - "reason": "Empty chat-template input crash versus XLMRobertaTokenizer passing a dict to Unigram; unrelated code paths." + "reason": "Empty chat-template list `IndexError` vs XLMRobertaTokenizer `Unigram(vocab=...)` type mismatch." }, { - "left": "issue:45593", - "right": "issue:47068", + "left": "issue:46132", + "right": "issue:47145", "accept": false, - "reason": "D-FINE auxiliary-loss omission is unrelated to the deepspeed host-synchronization concern." + "reason": "AttentionInterface registration behavior change vs FP8 checkpoint loading without weight_scale; different failures." }, { - "left": "issue:44483", + "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "Chat-completions API request rejection is not the same as Gemma pipeline token roundtripping." + "reason": "`torch.compile` attention-kernel divergence vs Gemma token roundtrip bug." }, { - "left": "issue:43976", - "right": "issue:47429", + "left": "issue:43784", + "right": "issue:46399", "accept": false, - "reason": "Python version compatibility report and tokenizers dependency upper bound are separate packaging issues." + "reason": "Missing `nn` import on import vs missing decorator for `apply_rotary_pos_emb` registration." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "resume_from_checkpoint key mismatch and data-filtering slowdown after upgrade are unrelated." + "reason": "`resume_from_checkpoint` key mismatch vs performance regression after upgrade." }, { - "left": "issue:44188", - "right": "issue:46421", + "left": "issue:43976", + "right": "issue:47429", "accept": false, - "reason": "Attention-kernel divergence under torch.compile is different from cache lazy-initialization recompilation." + "reason": "Python version support regression vs tokenizers upper-bound packaging constraint." }, { - "left": "issue:43784", - "right": "issue:46399", + "left": "issue:44188", + "right": "issue:46421", "accept": false, - "reason": "Import-time NameError in sentence-transformers is unrelated to Qwen3.5 kernel registration failure." + "reason": "Attention-kernel divergence under compile vs cache lazy initialization causing recompilation." }, { - "left": "issue:44749", - "right": "issue:45656", + "left": "issue:44483", + "right": "issue:46202", "accept": false, - "reason": "A slowdown regression in filtering data is not the same as deepspeed calling optimizer.step twice." + "reason": "`/v1/chat/completions` request rejection vs Gemma pipeline roundtrip failure; unrelated." }, { - "left": "issue:43065", - "right": "issue:46123", + "left": "issue:39692", + "right": "issue:46726", "accept": false, - "reason": "SAM3 pixel decoder dummy Conv2d and mask pipeline final-batch handling are unrelated." + "reason": "SigLIP2 docs example errors vs generic missing-vision-backend processor error; not the same bug." }, { - "left": "issue:39692", - "right": "issue:46726", + "left": "issue:44749", + "right": "issue:45656", "accept": false, - "reason": "SigLIP2 doc example errors and AutoProcessor backend-missing handling are different vision/documentation issues." + "reason": "Filtering slowdown after upgrade vs Deepspeed double optimizer step; different subsystems." } ] }, @@ -15261,6 +15338,7 @@ "issue:41628", "issue:42907", "issue:42994", + "issue:43065", "issue:43232", "issue:43475", "issue:43482", @@ -15280,18 +15358,17 @@ "issue:44661", "issue:44869", "issue:44871", - "issue:44933", "issue:44945", "issue:44964", "issue:45137", "issue:45230", - "issue:45310", "issue:45356", "issue:45375", "issue:45397", "issue:45478", "issue:45561", "issue:45656", + "issue:46123", "issue:46132", "issue:46326", "issue:46463", @@ -15305,237 +15382,224 @@ "issue:47448" ], "soft_pairs": [ - "issue:44297|issue:44623", - "issue:43475|issue:46821", "issue:42994|issue:43482", + "issue:43065|issue:46123", + "issue:43475|issue:46821", "issue:43475|issue:43643", "issue:43646|issue:44162", - "issue:44964|issue:45137", "issue:46463|issue:46877", - "issue:44373|issue:46851", + "issue:44297|issue:44623", + "issue:44964|issue:45137", "issue:42907|issue:47448", - "issue:45230|issue:45397", - "issue:44623|issue:45656", + "issue:44373|issue:46851", "issue:28282|issue:30990", - "issue:44871|issue:46132", "issue:43723|issue:46326", + "issue:44623|issue:45656", + "issue:45230|issue:45397", + "issue:44871|issue:46132", "issue:44188|issue:46851", + "issue:46877|issue:47000", + "issue:44451|issue:47429", "issue:45561|issue:47217", "issue:43232|issue:43673", - "issue:44451|issue:47429", - "issue:46877|issue:47000", - "issue:44279|issue:47429", - "issue:45375|issue:45478", "issue:44661|issue:47059", - "issue:44869|issue:45356", - "issue:44038|issue:44623", "issue:36010|issue:41628", + "issue:44869|issue:45356", + "issue:44279|issue:47429", "issue:43873|issue:44945", - "issue:45310|issue:45375", - "issue:41628|issue:44933" + "issue:45375|issue:45478", + "issue:44038|issue:44623" ], - "prepared_review_unit_hash": "453b97890e57e459c19c1e3a00eabb4190413fdab3daaa86db3c159ec4986def", + "prepared_review_unit_hash": "82ea9501d21103999cc9a86cdfd11442dd082f6a0a35f6a60cdf53138bedaf9e", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 35519, - "estimated_input_tokens": 8880, - "estimated_eval_tokens": 18016 + "soft_pair_count": 27, + "serialized_chars": 35254, + "estimated_input_tokens": 8814, + "estimated_eval_tokens": 17884 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose collection of distinct Transformers regressions (imports, loading, saving, generation, quantization, and model-specific failures). The soft pairs share broad wording but do not look like the same underlying bug/change, so I reject all of them.", - "confidence": 0.94, - "canonical_issue_reason": "issue:43646 is the broadest representative of the cluster\u2019s recurring v5 regression theme, but the set is still heterogeneous and not a clean duplicate group.", + "summary": "This is a mixed cluster of mostly unrelated Transformers issues, many around v5 loading/init regressions, tokenizer/model-specific failures, and a few unrelated utility/test problems. None of the soft pairs are strong enough to treat as duplicates under a strict code-path standard; the broadest representative issue is the generic v5 custom model initialization regression.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43646 is the broadest and most representative issue in the set: it describes a general Transformers 5.0.0 model-initialization break, which aligns with the cluster\u2019s recurring v5 compatibility theme better than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the most generally phrased issue and best fits as a cluster placeholder, though it only loosely represents the rest of the items.", + "best_issue_reason": "issue:43646 is the best issue to keep as the cluster representative because it is general, clearly user-facing, and likely to absorb the broadest class of v5 initialization regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44297", - "right": "issue:44623", + "left": "issue:42994", + "right": "issue:43482", "accept": false, - "reason": "Both are save/pretrained-related, but one is a tokenizer config mismatch and the other is missing processor files; different components and failure modes." + "reason": "One is about quantized model saving, the other about GGUF loading in v5; different operations and failure modes." }, { - "left": "issue:43475", - "right": "issue:46821", + "left": "issue:43065", + "right": "issue:46123", "accept": false, - "reason": "These are unrelated model-specific attribute errors in different code paths and different models." + "reason": "SAM3 pixel decoder internals vs MaskGenerationPipeline batch-finalization logic are unrelated code paths." }, { - "left": "issue:42994", - "right": "issue:43482", + "left": "issue:43475", + "right": "issue:46821", "accept": false, - "reason": "One is a quantized model save failure, the other a GGUF load failure on v5; opposite directions and different root problems." + "reason": "Different models and different missing-attribute errors; no shared concrete bug." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "Both mention missing fields/attributes, but one is a SAM3 model output bug and the other is AutoConfig with trust_remote_code; not the same bug." + "reason": "SAM3 vision encoder output attribute issue versus AutoConfig trust_remote_code field loss; distinct loading/serialization problems." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "Both are broad v5 regressions, but they affect different model families with different symptoms and no clear shared code path." - }, - { - "left": "issue:44964", - "right": "issue:45137", - "accept": false, - "reason": "A model loading regression and a DeepSpeed ZeRO3 IndexError are not the same underlying issue." + "reason": "Both are v5 regressions, but the concrete failures differ: custom model init versus ESM2-specific breakage." }, { "left": "issue:46463", "right": "issue:46877", "accept": false, - "reason": "An error-message bug in T5 validation and import-time docstring diagnostics are unrelated." + "reason": "An incorrect decoder-input error message is not the same bug as docstring diagnostics printing at import time." }, { - "left": "issue:44373", - "right": "issue:46851", + "left": "issue:44297", + "right": "issue:44623", "accept": false, - "reason": "One is a docstring typo, the other is a runtime variable-length position_ids bug; same term, different problem." + "reason": "Tokenizer save metadata mismatch and processor save missing files are different save_pretrained problems." }, { - "left": "issue:42907", - "right": "issue:47448", + "left": "issue:44964", + "right": "issue:45137", "accept": false, - "reason": "Different models and different failures: saving dequantized models versus architecture recognition." + "reason": "Phi-4 multimodal loading and DeepSpeed ZeRO3 deque errors are unrelated runtime failures." }, { - "left": "issue:45230", - "right": "issue:45397", + "left": "issue:42907", + "right": "issue:47448", "accept": false, - "reason": "The first is too generic to anchor a duplicate, and the second is a specific Gemma-4 Zero3 from_pretrained failure." + "reason": "Missing PyTorch dependency error versus Gemma4 architecture recognition failure are different import/load issues." }, { - "left": "issue:44623", - "right": "issue:45656", + "left": "issue:44373", + "right": "issue:46851", "accept": false, - "reason": "Processor save-pretrained file generation and optimizer stepping under DeepSpeed are unrelated." + "reason": "Wrong docstring text and variable-length position_ids runtime behavior are not the same bug." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "Both are loading-related user reports, but one is a missing PyTorch dependency error and the other is a hang while loading Sentence Transformers." - }, - { - "left": "issue:44871", - "right": "issue:46132", - "accept": false, - "reason": "Gemma eos_token_id mismatch and AttentionInterface.register behavior changes are unrelated subsystems." + "reason": "ImportError for missing PyTorch is unrelated to Sentence Transformers hanging during load." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "Both involve tokenizers, but one is AutoTokenizer loading in v5 and the other is continue_final_message on mistral_common tokenizers." + "reason": "AutoTokenizer loading in v5 and continue_final_message behavior for mistral_common are separate tokenizer-path issues." }, { - "left": "issue:44188", - "right": "issue:46851", + "left": "issue:44623", + "right": "issue:45656", "accept": false, - "reason": "Different generation-time bugs: attention kernel divergence under torch.compile versus variable-length position_ids handling." + "reason": "Missing processor save artifacts and duplicate optimizer steps under deepspeed are unrelated." }, { - "left": "issue:45561", - "right": "issue:47217", + "left": "issue:45230", + "right": "issue:45397", "accept": false, - "reason": "A pytest-xdist test-utils race and a VL image-placeholder model bug are unrelated." + "reason": "A generic bug report title does not establish the same concrete failure as the Gemma-4 Zero3 loading issue." }, { - "left": "issue:43232", - "right": "issue:43673", + "left": "issue:44871", + "right": "issue:46132", "accept": false, - "reason": "Both are generation internals, but one is sync_gpus/_update_model_kwargs and the other is chunked_prefill cache handling." + "reason": "Gemma-3 EOS-token mismatch and AttentionInterface.register behavior change are different model/config issues." }, { - "left": "issue:44451", - "right": "issue:47429", + "left": "issue:44188", + "right": "issue:46851", "accept": false, - "reason": "A model load failure and a tokenizers version upper-bound packaging issue are not the same bug." + "reason": "torch.compile attention-kernel divergence is a different code path than Qwen3.5 variable-length position_ids handling." }, { "left": "issue:46877", "right": "issue:47000", "accept": false, - "reason": "Import-time lint output and RTDetrLoss infeasible cost matrix are unrelated." + "reason": "Import-time doc-lint output and RTDetrLoss infeasible cost matrix are unrelated." }, { - "left": "issue:44279", + "left": "issue:44451", "right": "issue:47429", "accept": false, - "reason": "One is a generic dependency complaint; the other is a specific tokenizers pin/release-blocking packaging issue." + "reason": "Model loading failure and dependency upper-bound packaging issue are not the same underlying bug." }, { - "left": "issue:45375", - "right": "issue:45478", + "left": "issue:45561", + "right": "issue:47217", "accept": false, - "reason": "Both are Qwen3.5 MoE-related, but one is a missing config field being dropped and the other is a from_pretrained error; not enough evidence they are the same bug." + "reason": "pytest-xdist file-race in tests is unrelated to VL placeholder handling in user text." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and missing GenerationMixin cache in chunked_prefill are different generation bugs." }, { "left": "issue:44661", "right": "issue:47059", "accept": false, - "reason": "A model-registration failure and a CLI completion attribute error are completely separate." + "reason": "Tokenizer mapping-name edge case and huggingface_hub CLI completion regression are unrelated." }, { - "left": "issue:44869", - "right": "issue:45356", + "left": "issue:36010", + "right": "issue:41628", "accept": false, - "reason": "Both are tokenizer-related regressions, but they concern different models and different decoding/configuration paths." + "reason": "Missing GenerationMixin import and missing AutoImageProcessor import are different API export regressions." }, { - "left": "issue:44038", - "right": "issue:44623", + "left": "issue:44869", + "right": "issue:45356", "accept": false, - "reason": "A Qwen3-VL-Moe loading bug and a processor.save_pretrained file issue are different problem classes." + "reason": "Whisper trailing replacement-character crash and Kimi-K2.5 codec-handling regression are not clearly the same code-path." }, { - "left": "issue:36010", - "right": "issue:41628", + "left": "issue:44279", + "right": "issue:47429", "accept": false, - "reason": "Both are import errors, but for different symbols and different API surfaces; not the same underlying change." + "reason": "General dependency complaint versus a specific tokenizers version-constraint problem are different issues." }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization/offloading behavior and pipeline-parallel output correctness are different distributed inference issues." - }, - { - "left": "issue:45310", - "right": "issue:45375", - "accept": false, - "reason": "They are related by model family, but one is a general from_pretrained error and the other is a specific config-field omission." + "reason": "Quantization offloading behavior and pipeline-parallelism output divergence are separate runtime behaviors." }, { - "left": "issue:41628", - "right": "issue:44933", + "left": "issue:45375", + "right": "issue:45478", "accept": false, - "reason": "Both mention imports from image utilities, but they do not point to the same missing symbol or change." + "reason": "Both mention Qwen3.5 MoE, but one is a missing config field dropped by strict serialization and the other is a broader from_pretrained failure; the concrete bugs are not identical enough to merge." }, { - "left": "issue:44933", - "right": "issue:47059", + "left": "issue:44038", + "right": "issue:44623", "accept": false, - "reason": "A missing image_utils import and a huggingface_hub CLI completion failure are unrelated." + "reason": "Qwen3-VL-Moe compatibility with Transformers 5.0 and processor.save_pretrained missing files are different problems." } ] }, @@ -15592,6 +15656,7 @@ "issue:45103", "issue:45137", "issue:45310", + "issue:45375", "issue:45405", "issue:45406", "issue:45561", @@ -15610,243 +15675,251 @@ "soft_pairs": [ "issue:45081|issue:45103", "issue:43278|issue:45137", + "issue:45310|issue:45375", + "issue:41628|issue:44933", "issue:42371|issue:45092", "issue:29925|issue:30333", "issue:45406|issue:46202", "issue:46202|issue:46701", "issue:44857|issue:46877", - "issue:45593|issue:46877", + "issue:44743|issue:47068", "issue:43232|issue:44792", + "issue:45593|issue:46877", "issue:43784|issue:43824", "issue:42175|issue:44933", "issue:44038|issue:45310", "issue:41628|issue:44351", - "issue:44222|issue:45310", - "issue:45405|issue:46877", "issue:30333|issue:33290", - "issue:44743|issue:47068", + "issue:45405|issue:46877", "issue:42898|issue:44568", - "issue:45561|issue:47059", + "issue:44222|issue:45310", "issue:42175|issue:45712", + "issue:42898|issue:45812", + "issue:45561|issue:47059", "issue:44945|issue:45753", "issue:42831|issue:44222", - "issue:42898|issue:45812", "issue:44496|issue:45561", - "issue:30990|issue:35141", "issue:44843|issue:47085", - "issue:44568|issue:47085", "issue:44186|issue:47085", - "issue:43976|issue:44933", - "issue:43479|issue:46829" + "issue:30990|issue:35141", + "issue:44568|issue:47085", + "issue:43479|issue:46829", + "issue:43976|issue:44933" ], - "prepared_review_unit_hash": "59cf4122f6de5671c5ebf76f71b2fd1f76758679ab023535f3e110f3dcf2d032", + "prepared_review_unit_hash": "fdfc324ab3aee9eaffab03a1d153de7fb4ae33e3bd1f36489d55f31211a9f30f", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 36175, - "estimated_input_tokens": 9044, - "estimated_eval_tokens": 18344 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37364, + "estimated_input_tokens": 9341, + "estimated_eval_tokens": 18938 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is highly heterogeneous: most pairs are clearly different bugs or unrelated infrastructure issues, so there is no real duplicate cluster here.", - "confidence": 0.95, - "canonical_issue_reason": "issue:45081 is the strongest single issue in this set: it has a concrete, localized crash in `_patch_mistral_regex`, a clear repro path, and multiple inbound references.", + "summary": "This looks like a noisy, heterogeneous cluster rather than a real duplicate set: the soft pairs span unrelated tokenizer, import, training, generation, FP8, and packaging issues, so none should be merged. If an anchor is needed, 47085 is the most substantive standalone report, but it does not unify the cluster.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out because the items are mostly unrelated false-positive similarities across different subsystems and failure modes.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45081 is the best representative issue overall because it is specific, actionable, and clearly scoped; the other items are not close duplicates of it.", + "best_issue_reason": "Issue 47085 is the clearest high-signal standalone bug report in the set: it has a concrete failure mode, discussion, and inbound references. It is not a cluster-wide duplicate target, just the strongest single issue here.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45081", "right": "issue:45103", "accept": false, - "reason": "Different subsystems and failures: Mistral regex crash vs auto_docstring annotation handling." + "reason": "Different failures: Mistral tokenizer regex handling vs docstring annotation parsing. No shared code path." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Unrelated bugs: dtype mismatch in evaluation vs DeepSpeed ZeRO3 deque error." + "reason": "Unrelated symptoms: embedding dtype drift in evaluation vs a DeepSpeed ZeRO3 deque error." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Both involve Qwen3.5-MoE, but one is a general from_pretrained error and the other is a missing vision-config field. Not the same bug." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they target different symbols and code paths." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "Different problems: TF backend packaging vs meta-init incompatibility for old InternVL2 checkpoints." + "reason": "Package/backend install behavior vs meta-initialization incompatibility for old InternVL2 checkpoints." }, { "left": "issue:29925", "right": "issue:30333", "accept": false, - "reason": "Tokenizer fast-mode enforcement is unrelated to MLflow reporting after failed training jobs." + "reason": "Tokenizer fast-mode enforcement and MLflow reporting are unrelated." }, { "left": "issue:45406", "right": "issue:46202", "accept": false, - "reason": "Processor `_tokenizer` attribute crash is not the same as Gemma roundtrip tokenization failure." + "reason": "Gemma processor attribute crash vs Gemma pipeline roundtrip failure; different concrete defects." }, { "left": "issue:46202", "right": "issue:46701", "accept": false, - "reason": "Both involve Gemma, but one is a token roundtrip issue and the other is an image-text pipeline failure." + "reason": "Both are Gemma-related, but one is token roundtrip and the other is an image-text pipeline/output-type issue." }, { "left": "issue:44857", "right": "issue:46877", "accept": false, - "reason": "CUDA float16 loss crash is unrelated to auto_docstring printing diagnostics at import time." + "reason": "CUDA/AMP crash in a loss module is unrelated to auto_docstring import-time stdout noise." }, { - "left": "issue:45593", - "right": "issue:46877", + "left": "issue:44743", + "right": "issue:47068", "accept": false, - "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring stdout diagnostics." + "reason": "Qwen recurrent-state reset bug and a CUDA host-synchronization concern in deepspeed_sp_compute_loss are different issues." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation kwargs sync bug and Janus image-generation test failure are different code paths." + "reason": "Generation kwargs update bug vs Janus image generation test failure; not the same code path." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring diagnostics printing." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Sentence-transformers import NameError is not the same as missing Qwen2.5-VL export." + "reason": "Separate import failures for different packages/models; no common underlying bug." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Packaging backend omission is unrelated to a missing image_utils import." + "reason": "Backend extras packaging issue vs missing image_utils export; different failures." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Both mention Qwen Moe models, but they describe different model/version-specific failures." + "reason": "Both mention Qwen MoE models, but they are different model families and different failure surfaces." }, { "left": "issue:41628", "right": "issue:44351", "accept": false, - "reason": "Missing AutoImageProcessor and missing HybridCache are separate export/import regressions." + "reason": "Import errors for different public symbols; not the same underlying regression." }, { - "left": "issue:44222", - "right": "issue:45310", + "left": "issue:30333", + "right": "issue:33290", "accept": false, - "reason": "FP8 save_pretrained behavior is not the same as Qwen3.5 Moe from_pretrained failure." + "reason": "MLflow failure reporting and AdaFactor OOM in DeepSpeed are unrelated." }, { "left": "issue:45405", "right": "issue:46877", "accept": false, - "reason": "PEFT version pinning is unrelated to auto_docstring import-time stderr/stdout output." + "reason": "Dependency version bump/release timing issue vs docstring lint output issue." }, { - "left": "issue:30333", - "right": "issue:33290", + "left": "issue:42898", + "right": "issue:44568", "accept": false, - "reason": "MLflow reporting and Adafactor OOM in DeepSpeed are unrelated failures." + "reason": "Both are tokenizer regressions in v5, but one is cleanup-space behavior and the other is add_special_tokens/BOS-EOS handling." }, { - "left": "issue:44743", - "right": "issue:47068", + "left": "issue:44222", + "right": "issue:45310", "accept": false, - "reason": "Recurrent-state reset in modular Qwen3.5 is different from a host-sync issue in deepspeed_sp_compute_loss." + "reason": "FP8 save_pretrained behavior and Qwen3.5 from_pretrained loading are different code paths." }, { - "left": "issue:42898", - "right": "issue:44568", + "left": "issue:42175", + "right": "issue:45712", "accept": false, - "reason": "Both are tokenizer regressions, but one is about cleanup-space behavior and the other about missing BOS/EOS insertion." + "reason": "Packaging/backend extras and leftover dummy PT classes are unrelated." }, { - "left": "issue:45561", - "right": "issue:47059", + "left": "issue:42898", + "right": "issue:45812", "accept": false, - "reason": "pytest-xdist file race is unrelated to the huggingface_hub CLI completion API break." + "reason": "Different tokenizer regressions: cleanup-space behavior vs Granite token ID generation." }, { - "left": "issue:42175", - "right": "issue:45712", + "left": "issue:45561", + "right": "issue:47059", "accept": false, - "reason": "TF backend installation and leftover dummy PT classes are different maintenance issues." + "reason": "A pytest-xdist file race and a huggingface_hub CLI compatibility break are unrelated." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallelism output corruption is unrelated to Qwen3_5 looping on a specific image." + "reason": "Pipeline parallelism output corruption and a Qwen3_5 image infinite loop are different failures." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FineGrainedFP8 accuracy and FP8 save_pretrained are different FP8-related bugs, not the same change." - }, - { - "left": "issue:42898", - "right": "issue:45812", - "accept": false, - "reason": "Different tokenizer regressions affecting different model families and different behaviors." + "reason": "FP8 accuracy issue vs FP8 save_pretrained issue; same broad area, not the same bug." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Model config auto-detection failure is unrelated to a pytest-xdist race condition." + "reason": "Unrecognized model/config problem vs xdist captured_info.txt race; unrelated." }, { - "left": "issue:30990", - "right": "issue:35141", + "left": "issue:44843", + "right": "issue:47085", "accept": false, - "reason": "Loading hang for sentence-transformers and embedding reinitialization in post_init are distinct issues." + "reason": "Offline mistral-regex patch bug and tokenizer thread-safety regression are different code paths and symptoms." }, { - "left": "issue:44843", + "left": "issue:44186", "right": "issue:47085", "accept": false, - "reason": "Offline model_info call in Mistral regex patch is unrelated to tokenizer thread-safety." + "reason": "LayoutLMv2Tokenizer batching crash is not the same as the fast-tokenizer thread-safety regression." }, { "left": "issue:44568", "right": "issue:47085", "accept": false, - "reason": "BOS/EOS token regression is not the same bug as Rust tokenizer thread-safety." + "reason": "BOS/EOS insertion regression and tokenizer thread-safety are distinct issues." }, { - "left": "issue:44186", - "right": "issue:47085", + "left": "issue:43479", + "right": "issue:46829", "accept": false, - "reason": "LayoutLMv2Tokenizer padding/NER crashes are different from a general tokenizer thread-safety regression." + "reason": "Config default initialization bug vs output field misassignment in VideoPrism; unrelated." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python version incompatibility in Transformers 5.1.0 is unrelated to a missing image_utils import." - }, - { - "left": "issue:43479", - "right": "issue:46829", - "accept": false, - "reason": "Phi4 multimodal config defaults and VideoPrism output-field mismatch are unrelated model bugs." + "reason": "Python-version compatibility break and missing image_utils import are unrelated." } ] }, @@ -15890,9 +15963,7 @@ "issue:44060", "issue:44079", "issue:44190", - "issue:44222", "issue:44295", - "issue:44297", "issue:44448", "issue:44568", "issue:44623", @@ -15914,243 +15985,244 @@ "issue:46209", "issue:46231", "issue:46431", + "issue:46665", "issue:46693", "issue:46909", "issue:47059", "issue:47085" ], "soft_pairs": [ - "issue:44060|issue:45125", "issue:42673|issue:46693", - "issue:42673|issue:43704", "issue:39692|issue:46909", - "issue:44623|issue:45310", + "issue:44060|issue:45125", "issue:43723|issue:44568", + "issue:42673|issue:43704", "issue:43479|issue:44079", - "issue:45600|issue:47059", + "issue:44623|issue:45310", "issue:30333|issue:36331", "issue:42898|issue:44448", - "issue:30333|issue:34689", "issue:42898|issue:43295", + "issue:45600|issue:47059", + "issue:46209|issue:46431", + "issue:30333|issue:34689", "issue:42673|issue:47085", "issue:42898|issue:44779", "issue:43479|issue:44869", + "issue:22355|issue:28282", "issue:44016|issue:46231", "issue:46077|issue:46123", "issue:42222|issue:45200", - "issue:22355|issue:28282", - "issue:46209|issue:46431", - "issue:44295|issue:45656", "issue:44734|issue:45103", "issue:42796|issue:43723", - "issue:43874|issue:45092", "issue:44448|issue:46132", + "issue:44295|issue:45656", "issue:43116|issue:44190", "issue:45405|issue:45712", - "issue:44222|issue:44297" + "issue:43874|issue:45092", + "issue:43479|issue:46665" ], - "prepared_review_unit_hash": "4385d7da9c8434ad8f12be0991eb1dda2346f638cbeeca227e4edeb391272173", + "prepared_review_unit_hash": "b72e4cc35fb5dae4798fd4e531bf376def76643253eb1a100fabc954e88ae98a", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 36280, - "estimated_input_tokens": 9070, - "estimated_eval_tokens": 18396 + "serialized_chars": 35723, + "estimated_input_tokens": 8931, + "estimated_eval_tokens": 18118 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of unrelated transformer issues. The only clear duplicate-style match is the repeated Qwen3ForCausalLM VRAM leak report; the other soft pairs share broad themes (tokenizers, docs, multimodal, v5 regressions) but not the same concrete bug.", - "confidence": 0.95, - "canonical_issue_reason": "issue:42673 is the best canonical issue because it is the earlier, more detailed report for the exact Qwen3ForCausalLM VRAM leak and has a direct duplicate-looking follow-up in issue:43704.", + "summary": "This set is mostly a heterogeneous mix of unrelated transformer bugs/docs regressions. The only clear duplicate-looking pair is the repeated Qwen3ForCausalLM VRAM leak report; the rest are merely same-subsystem or same-version similarity, not the same underlying issue.", + "confidence": 0.89, + "canonical_issue_reason": "issue:42673 is the best canonical issue because it has an exact duplicate in issue:43704 and describes a concrete, specific bug with a clear symptom and reproduction scope.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42673 is the strongest representative issue overall: it is specific, clearly described, and has the clearest duplicate counterpart in the set.", + "best_issue_reason": "issue:42673 is the strongest issue to keep as the representative anchor: it is specific, duplicated verbatim by another report, and more concrete than the broader version/regression reports in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44060", - "right": "issue:45125", - "accept": false, - "reason": "Both are Qwen/Qwen3-family reports, but one is a tied-weights warning and the other is a missing tensor-parallel plan; different bugs and code paths." - }, { "left": "issue:42673", "right": "issue:46693", "accept": false, - "reason": "VRAM leak vs FA2 long-context performance regression; different symptoms and likely different subsystems." - }, - { - "left": "issue:42673", - "right": "issue:43704", - "accept": true, - "reason": "Same title and same underlying Qwen3ForCausalLM VRAM leak; this looks like a duplicate/re-report of the same bug." + "reason": "Both mention performance/VRAM around Qwen3, but one is a dataloader-thread leak and the other is an FA2 long-context performance regression; different bug paths." }, { "left": "issue:39692", "right": "issue:46909", "accept": false, - "reason": "Both are broken docs examples, but for different models and different failure modes." + "reason": "Both are broken documentation examples, but they target different models and failures (SigLIP2 example vs DIA example), so not the same issue." }, { - "left": "issue:44623", - "right": "issue:45310", + "left": "issue:44060", + "right": "issue:45125", "accept": false, - "reason": "Processor save_pretrained missing files vs Qwen3.5-Moe from_pretrained error; unrelated code paths." + "reason": "Both involve Qwen3 model internals, but one is a tied-weights warning and the other is a missing tensor-parallel plan; distinct code-path problems." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Tokenizer loading regression vs special-token insertion regression; both tokenizer-related but not the same concrete issue." + "reason": "Both are tokenizer/v5 regressions, but one is tokenizer loading and the other is add_special_tokens behavior for a specific model; not the same bug." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same Qwen3ForCausalLM VRAM leak symptom; these look like duplicate reports of the same bug." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Multimodal config defaults vs ModelOutput key assignment; unrelated bugs." + "reason": "One is multimodal config initialization defaults, the other is ModelOutput key assignment when a key was previously None; unrelated internals." }, { - "left": "issue:45600", - "right": "issue:47059", + "left": "issue:44623", + "right": "issue:45310", "accept": false, - "reason": "auto_mappings CI breakage vs CLI completion API breakage; different areas entirely." + "reason": "Different failures: missing files in processor.save_pretrained versus a Qwen3.5 MoE from_pretrained error." }, { "left": "issue:30333", "right": "issue:36331", "accept": false, - "reason": "MLflow reporting on failed training vs CustomTrainer compute_loss signature mismatch; not the same problem." + "reason": "MLflow reporting on failed training jobs is unrelated to CustomTrainer.compute_loss receiving an unexpected keyword argument." }, { "left": "issue:42898", "right": "issue:44448", "accept": false, - "reason": "General v5 tokenization cleanup behavior change vs Pegasus output difference; related theme, but not clearly the same bug." + "reason": "Both are v5 behavior regressions, but one is tokenization cleanup and the other is Pegasus output drift; different symptoms and fixes." }, { - "left": "issue:30333", - "right": "issue:34689", + "left": "issue:42898", + "right": "issue:43295", "accept": false, - "reason": "Training-job reporting issue vs model loading regression for Llama 3.2 vision; unrelated." + "reason": "The first is about tokenization cleanup behavior; the second is a processor/tokenizer API regression with image inputs. Not the same issue." }, { - "left": "issue:42898", - "right": "issue:43295", + "left": "issue:45600", + "right": "issue:47059", + "accept": false, + "reason": "CI breakage from removed config references is unrelated to the huggingface_hub CLI completion failure." + }, + { + "left": "issue:46209", + "right": "issue:46431", "accept": false, - "reason": "Both are v5 behavior regressions, but one is clean-up spacing semantics and the other is processor/tokenizer access with images." + "reason": "FP8 inference on B200 and missing chat template are separate runtime problems with different code paths." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "MLflow job-status reporting and Llama 3.2 Vision model loading are unrelated bugs." }, { "left": "issue:42673", "right": "issue:47085", "accept": false, - "reason": "Both mention threading and tokenizers/models, but one is a specific VRAM leak in Qwen3ForCausalLM and the other is a broad thread-safety regression." + "reason": "A VRAM leak in Qwen3ForCausalLM is not the same as a broader tokenizer thread-safety regression." }, { "left": "issue:42898", "right": "issue:44779", "accept": false, - "reason": "Tokenizer cleanup behavior change vs DeepSeek tokenizer correctness regression; different concrete failures." + "reason": "Both are tokenizer regressions in v5, but they affect different models and behaviors; not mergeable as one bug fix." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Multimodal config initialization vs Whisper timestamp decoding crash; unrelated." + "reason": "Multimodal config defaults and Whisper timestamp decoding are unrelated failures." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Different import errors: missing transformers.onnx versus missing PyTorch for AutoModel." }, { "left": "issue:44016", "right": "issue:46231", "accept": false, - "reason": "Notebook syntax error vs documentation parameter-name mismatch; different documentation issues." + "reason": "A notebook syntax error and a doc parameter-name mismatch are separate documentation issues." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Bidirectional attention masking bug vs partial-batch pipeline bookkeeping bug; different code paths." + "reason": "Bidirectional attention mask construction and final partial-batch result dropping are different pipeline/model behaviors." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "VitPose model breakage vs Gemma 4 multimodal token-type defaulting; different model families and failures." - }, - { - "left": "issue:22355", - "right": "issue:28282", - "accept": false, - "reason": "Different import errors: missing transformers.onnx module vs missing PyTorch dependency." - }, - { - "left": "issue:46209", - "right": "issue:46431", - "accept": false, - "reason": "FP8 inference failure vs missing chat template on processor; unrelated." - }, - { - "left": "issue:44295", - "right": "issue:45656", - "accept": false, - "reason": "position_ids buffer/read bug vs deepspeed optimizer double-step; unrelated." + "reason": "Vitpose model breakage and Gemma 4 mm_token_type_ids defaulting are distinct model-family bugs." }, { "left": "issue:44734", "right": "issue:45103", "accept": false, - "reason": "KV-cache continuation tensor indexing vs auto_docstring future-annotations crash; unrelated." + "reason": "Serving KV-cache indexing and auto_docstring annotation handling are unrelated code paths." }, { "left": "issue:42796", "right": "issue:43723", "accept": false, - "reason": "Tokenizer BPE-marker handling vs AutoTokenizer.from_pretrained loading issue; similar area, but not the same bug." + "reason": "Both involve tokenizer-related v5 behavior, but one is BPE marker handling for Ministral and the other is AutoTokenizer loading; different defects." }, { - "left": "issue:43874", - "right": "issue:45092", + "left": "issue:44448", + "right": "issue:46132", "accept": false, - "reason": "Image processor method missing vs remote-code checkpoint/meta-init incompatibility; different multimodal issues." + "reason": "Pegasus output differences and AttentionInterface.register behavior changes are not the same underlying bug." }, { - "left": "issue:44448", - "right": "issue:46132", + "left": "issue:44295", + "right": "issue:45656", "accept": false, - "reason": "Model output differences across v4/v5 vs AttentionInterface.register behavior change; not the same issue." + "reason": "A position_ids buffer/read error and a double optimizer step under deepspeed are unrelated training/runtime bugs." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Two example-script failures, but one is multi-label classification output and the other is local dataset loading; not duplicates." + "reason": "Different example scripts and different failures: empty multilabel outputs versus loading a local image dataset." }, { "left": "issue:45405", "right": "issue:45712", "accept": false, - "reason": "PEFT version pin issue vs leftover dummy classes in repo; unrelated." + "reason": "A version pin mismatch for peft is unrelated to leftover dummy class cleanup." }, { - "left": "issue:44222", - "right": "issue:44297", + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-processor method and remote-code meta-initialization incompatibility are separate multimodal-loading problems." + }, + { + "left": "issue:43479", + "right": "issue:46665", "accept": false, - "reason": "FP8 save_pretrained moe vs tokenizer save_pretrained class-name mismatch; different subsystems." + "reason": "Multimodal config default initialization is unrelated to inconsistent from_pretrained behavior inside Python classes." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet: issue:42673 and issue:43704 are clearly the same Qwen3ForCausalLM VRAM leak by title and chronology, while the other pairs are conservatively marked non-duplicates based on distinct titles/subsystems. No evidence is overstated." + "feedback": "The analyst output is grounded in the packet and stays conservative on the soft-edge matches. The only clearly duplicate-looking pair is issue:42673 and issue:43704, which share the exact title and the same Qwen3ForCausalLM VRAM leak symptom. The rest of the pairwise rejections appropriately distinguish different bug paths or failures rather than overstating similarity." }, "evaluator_used": true, "retried": false, @@ -16176,19 +16248,19 @@ "issue:41628", "issue:42757", "issue:42831", + "issue:42907", "issue:43122", "issue:43232", "issue:43408", "issue:43475", - "issue:43479", "issue:43504", "issue:43723", + "issue:43986", "issue:44075", "issue:44190", "issue:44222", "issue:44297", "issue:44327", - "issue:44360", "issue:44373", "issue:44484", "issue:44485", @@ -16200,9 +16272,11 @@ "issue:45245", "issue:45310", "issue:45357", + "issue:45381", "issue:45656", "issue:45712", "issue:45834", + "issue:46050", "issue:46082", "issue:46132", "issue:46209", @@ -16212,239 +16286,255 @@ "issue:46402", "issue:46424", "issue:46614", - "issue:46665", - "issue:46762", + "issue:46726", "issue:46909", "issue:46932", "issue:47059", "issue:47276" ], "soft_pairs": [ - "issue:43479|issue:46665", + "issue:44222|issue:44297", "issue:42831|issue:44623", "issue:44297|issue:45656", - "issue:45245|issue:46132", "issue:44933|issue:45712", - "issue:44222|issue:44623", + "issue:45245|issue:46132", "issue:44075|issue:46231", - "issue:44373|issue:44485", + "issue:44222|issue:44623", "issue:43723|issue:44977", "issue:43232|issue:47276", - "issue:44297|issue:45310", + "issue:44373|issue:44485", "issue:43504|issue:47059", - "issue:44327|issue:47059", + "issue:44297|issue:45310", "issue:43122|issue:46209", - "issue:43475|issue:46305", - "issue:46614|issue:46932", "issue:43232|issue:45834", + "issue:46614|issue:46932", + "issue:43475|issue:46305", + "issue:44327|issue:47059", + "issue:30990|issue:38175", "issue:41628|issue:42757", "issue:45071|issue:46082", - "issue:30990|issue:38175", "issue:45092|issue:45357", + "issue:43408|issue:46402", "issue:44190|issue:46082", - "issue:44222|issue:44484", + "issue:42907|issue:46050", "issue:43232|issue:46424", - "issue:43408|issue:46402", "issue:46369|issue:46909", - "issue:44360|issue:46762" + "issue:45381|issue:46305", + "issue:43986|issue:46726", + "issue:44222|issue:44484" ], - "prepared_review_unit_hash": "d3c3203e8544280b0a305b6c308479129a0147e202cc842d0a82201229db9670", + "prepared_review_unit_hash": "928c426eaffaa4b2d4e421737e93252c0b43078f3891504832ee6c760c8a3915", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 34871, - "estimated_input_tokens": 8718, - "estimated_eval_tokens": 17692 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36068, + "estimated_input_tokens": 9017, + "estimated_eval_tokens": 18290 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: it contains unrelated issues across loading, saving, config validation, generation, tokenizer, FP8, and model-specific regressions. None of the proposed soft pairs look like the same underlying bug or change, so I rejected all of them.", - "confidence": 0.46, - "canonical_issue_reason": "I\u2019d anchor on issue:43479 because it is a concrete, self-contained bug report with a clear failure mode and fix path, making it the most representative standalone issue in this mixed set.", + "summary": "The set is mostly heterogeneous with many title-level false positives. I see one plausible duplicate thread around AutoProcessor failures when a required vision backend is missing; everything else looks like distinct bugs or unrelated documentation/config issues.", + "confidence": 0.38, + "canonical_issue_reason": "No single issue cleanly represents the whole set. If one representative must be chosen, issue 46726 is the broadest and most general description of the only plausible duplicate thread (AutoProcessor/backend-missing error handling).", "canonical_pr_reason": null, - "best_issue_reason": "Issue:43723 is the broadest user-facing loading regression in the set and the best single issue to represent the cluster if one must be chosen.", + "best_issue_reason": "Issue 46726 is the clearest, most general formulation of the AutoProcessor missing-vision-backend failure and can subsume the more specific report in 43986.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43479", - "right": "issue:46665", + "left": "issue:44222", + "right": "issue:44297", "accept": false, - "reason": "Both involve loading behavior, but one is Phi4Multimodal config defaults and the other is from_pretrained inside Python classes; different bug paths." + "reason": "Both mention save_pretrained, but one is FP8 model serialization and the other is tokenizer metadata; different bugs." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FineGrainedFP8 accuracy and processor.save_pretrained missing files are unrelated failures." + "reason": "FineGrainedFP8 accuracy vs missing processor.save_pretrained files are unrelated." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer save metadata mismatch vs optimizer step called twice; no shared code-path bug." - }, - { - "left": "issue:45245", - "right": "issue:46132", - "accept": false, - "reason": "Category-limit runtime error and AttentionInterface.register behavior change are distinct issues." + "reason": "Tokenizer save metadata and optimizer step duplication are different code paths." }, { "left": "issue:44933", "right": "issue:45712", "accept": false, - "reason": "Both touch imports, but one is a missing image_utils import and the other is leftover dummy classes leaking into dir/checks." + "reason": "Both touch import/export surfaces, but one is a nonexistent image_utils import and the other is dummy class cleanup/leakage; not the same defect." }, { - "left": "issue:44222", - "right": "issue:44623", + "left": "issue:45245", + "right": "issue:46132", "accept": false, - "reason": "Both mention save_pretrained, but one is FP8 MoE serialization and the other is processor file output; different artifacts and bugs." + "reason": "Category-count validation and AttentionInterface behavior changes are unrelated." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "Optimizer args ignored is a runtime bug; the other is just a doc/parameter-name mismatch." + "reason": "Optimizer argument handling and a training docstring typo are not the same issue." }, { - "left": "issue:44373", - "right": "issue:44485", + "left": "issue:44222", + "right": "issue:44623", "accept": false, - "reason": "Wrong docstring for position_ids is unrelated to GLM-5 RoPE implementation." + "reason": "FP8 model save_pretrained failure vs processor.save_pretrained missing files are different serialization problems." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "Tokenizer loading regression and Qwen3.5 flash-attention generation failure are unrelated." + "reason": "Tokenizer loading in v5 and Qwen3.5 flash-attention generation are unrelated." }, { "left": "issue:43232", "right": "issue:47276", "accept": false, - "reason": "One is a generation-state bug after sync_gpus; the other is only a misleading warning message." + "reason": "Both involve synced_gpus, but one is a generation kwargs/update bug and the other is a warning message issue." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "A docstring mismatch and a RoPE implementation discussion are not the same bug." + }, + { + "left": "issue:43504", + "right": "issue:47059", + "accept": false, + "reason": "Pretrained model preset loading and a huggingface_hub CLI compatibility failure are unrelated." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer_class mismatch on save and Qwen3.5 MoE from_pretrained errors are different issues." + "reason": "Tokenizer save metadata mismatch and Qwen3.5 from_pretrained failure are different problems." }, { - "left": "issue:43504", - "right": "issue:47059", + "left": "issue:43122", + "right": "issue:46209", "accept": false, - "reason": "BEiT legacy-field loading bug and a CLI incompatibility with huggingface_hub are unrelated." + "reason": "Tokenization drift across versions and FP8 inference failure on B200 are unrelated." }, { - "left": "issue:44327", - "right": "issue:47059", + "left": "issue:43232", + "right": "issue:45834", "accept": false, - "reason": "QA decode_spans top-k crash is a model pipeline bug; the hub error is an external CLI/API incompatibility." + "reason": "Generation state sync and a Kosmos2.5 OCR index error are different bugs." }, { - "left": "issue:43122", - "right": "issue:46209", + "left": "issue:46614", + "right": "issue:46932", "accept": false, - "reason": "Tokenizer output changes across versions and FP8 inference failing on B200 are different problems." + "reason": "Stale phonemizer language state and Gemma4 quantized embedder dtype casting are unrelated." }, { "left": "issue:43475", "right": "issue:46305", "accept": false, - "reason": "SAM3 attribute error and MoonshineStreaming off-by-one mask logic are unrelated." + "reason": "SAM3 attribute error and Moonshine streaming mask off-by-one are different code paths." }, { - "left": "issue:46614", - "right": "issue:46932", + "left": "issue:44327", + "right": "issue:47059", "accept": false, - "reason": "Stale phonemizer language state and Gemma4 dtype casting are different subsystems and failure modes." + "reason": "QA decode_spans bounds error and CLI completion API breakage are unrelated." }, { - "left": "issue:43232", - "right": "issue:45834", + "left": "issue:30990", + "right": "issue:38175", "accept": false, - "reason": "Generation kwargs update bug and Kosmos2.5 long-OCR index error are unrelated." + "reason": "Sentence Transformers loading hang and zero-probability SigLIP outputs are different issues." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they concern different symbols, modules, and root causes." + "reason": "Both are import errors, but for different missing symbols in different packages." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "PretrainedConfig type-checking regression and LlamaConfig head_dim validation are distinct config issues." + "reason": "PretrainedConfig type checking and LlamaConfig head_dim validation are related only superficially." }, { - "left": "issue:30990", - "right": "issue:38175", + "left": "issue:45092", + "right": "issue:45357", "accept": false, - "reason": "Stuck loading Sentence Transformers and zero probabilities in SigLIP2 are unrelated symptoms." + "reason": "Both involve model loading/saving, but the affected models and failure modes differ." }, { - "left": "issue:45092", - "right": "issue:45357", + "left": "issue:43408", + "right": "issue:46402", "accept": false, - "reason": "Remote-code meta-init incompatibility and incorrect visual encoder keys on save are different Qwen/InternVL multimodal bugs." + "reason": "sam3_video/sam3_tracker warning and missing Sam3Model text encoder weights are different SAM3 issues." }, { "left": "issue:44190", "right": "issue:46082", "accept": false, - "reason": "Local dataset loading script failure and LlamaConfig head_dim rejection are unrelated." + "reason": "Local dataset loading and LlamaConfig validation are unrelated." }, { - "left": "issue:44222", - "right": "issue:44484", + "left": "issue:42907", + "right": "issue:46050", "accept": false, - "reason": "FP8 MoE save_pretrained and 50GB shard-size default are different save-path concerns." + "reason": "Both are quantization save/load problems, but dequantized Ministrals/Devstrals and sinq serialization are different mechanisms." }, { "left": "issue:43232", "right": "issue:46424", "accept": false, - "reason": "Generation kwargs handling and cache max_cache_length behavior are separate cache/generation bugs." + "reason": "Generation sync_gpus handling and cache length behavior are unrelated." }, { - "left": "issue:43408", - "right": "issue:46402", + "left": "issue:46369", + "right": "issue:46909", "accept": false, - "reason": "SAM3 model-type warning and missing text encoder weights are not the same bug." + "reason": "Memory overflow in sam3-litetext and a broken documentation example are unrelated." }, { - "left": "issue:46369", - "right": "issue:46909", + "left": "issue:45381", + "right": "issue:46305", "accept": false, - "reason": "Model memory overflow and a broken documentation example are unrelated." + "reason": "Qwen video position_ids and Moonshine streaming mask off-by-one are different model-specific bugs." }, { - "left": "issue:44360", - "right": "issue:46762", + "left": "issue:43986", + "right": "issue:46726", + "accept": true, + "reason": "Both describe AutoProcessor failing when a required vision backend is missing; the same error-handling fix could plausibly cover both." + }, + { + "left": "issue:44222", + "right": "issue:44484", "accept": false, - "reason": "DSA indexer ReLU discussion and Minimax attention numerical differences are different math/code paths." + "reason": "FP8 model save_pretrained and max_shard_size default behavior are unrelated." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted soft edge, issue:43986 vs issue:46726, is plausibly the same AutoProcessor/vision-backend error-handling problem. The other pairwise rejects are supported by the titles and do not overclaim duplicates." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -16468,7 +16558,9 @@ "issue:41628", "issue:42796", "issue:42831", - "issue:42907", + "issue:42886", + "issue:42994", + "issue:43066", "issue:43122", "issue:43408", "issue:43540", @@ -16499,256 +16591,255 @@ "issue:45600", "issue:45753", "issue:46002", - "issue:46050", "issue:46305", "issue:46328", "issue:46369", "issue:46655", - "issue:46726", + "issue:46762", "issue:46909", "issue:47059", "issue:47217" ], "soft_pairs": [ - "issue:42907|issue:46050", - "issue:44263|issue:45092", "issue:44295|issue:44623", - "issue:43986|issue:46726", - "issue:45381|issue:46305", - "issue:44263|issue:44373", "issue:44279|issue:46328", + "issue:44263|issue:45092", + "issue:44360|issue:46762", + "issue:41628|issue:43824", + "issue:44263|issue:44373", "issue:46305|issue:46655", + "issue:22355|issue:28218", "issue:22355|issue:27957", "issue:43784|issue:45600", - "issue:41628|issue:43824", - "issue:22355|issue:28218", "issue:44483|issue:45406", + "issue:43756|issue:44060", "issue:42831|issue:45310", - "issue:43576|issue:43784", "issue:44964|issue:47059", - "issue:43756|issue:44060", - "issue:44261|issue:45092", - "issue:44263|issue:44360", + "issue:43576|issue:43784", "issue:44188|issue:44373", - "issue:42831|issue:44038", - "issue:44829|issue:46002", - "issue:43986|issue:47217", + "issue:44261|issue:45092", "issue:42796|issue:43122", + "issue:42831|issue:44038", + "issue:44263|issue:44360", "issue:43644|issue:46305", - "issue:42831|issue:45753", - "issue:39401|issue:44568", + "issue:43986|issue:47217", + "issue:44829|issue:46002", "issue:43540|issue:46909", + "issue:39401|issue:44568", + "issue:42831|issue:45753", "issue:43408|issue:46369", - "issue:45381|issue:46909" + "issue:45381|issue:46909", + "issue:42886|issue:43066", + "issue:42886|issue:42994" ], - "prepared_review_unit_hash": "84b3ea107489445d9928bc8bb3488b34771d83dc9ea5e0715814680273ea2346", + "prepared_review_unit_hash": "07765c1b7523c4bfa869cda36b82a52222b5399aca4150d5733e031367abb368", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 35886, - "estimated_input_tokens": 8972, - "estimated_eval_tokens": 18200 + "serialized_chars": 36387, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pairs look like true duplicates. The set is a grab bag of unrelated import, tokenizer, model-loading, processor, training, and documentation issues, so there is no clear canonical issue or PR.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This cluster is a loose mix of Transformers issues, mostly tokenizer/version-regression reports plus several unrelated model, import, and processor bugs. None of the soft pairs look like safe duplicates, so they should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43122 is the best available anchor because it is the broadest tokenizer regression report (same tokenizer, different behavior across 4.57.3 to 5.0) and sits near several other tokenizer-related issues. That said, the cluster is still too heterogeneous for a single true duplicate canon.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:43122 is the strongest representative issue among this set because it captures the common tokenizer/version-regression theme better than the more model-specific reports. It is still only a partial fit for the overall cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42907", - "right": "issue:46050", + "left": "issue:44295", + "right": "issue:44623", "accept": false, - "reason": "Different quantization/saving flows: dequantized Ministral/Devstral persistence vs sinq quantized reload behavior." + "reason": "Buffer read error and processor.save_pretrained missing files are different failure modes in different code paths." }, { - "left": "issue:44263", - "right": "issue:45092", + "left": "issue:44279", + "right": "issue:46328", "accept": false, - "reason": "DSA indexer split semantics vs InternVL2 meta-init compatibility are unrelated bugs." + "reason": "A dependency issue and a padding performance regression are unrelated." }, { - "left": "issue:44295", - "right": "issue:44623", + "left": "issue:44263", + "right": "issue:45092", "accept": false, - "reason": "Buffer read error for position_ids is not the same as processor.save_pretrained() missing files." + "reason": "torch.split return-value handling in GlmMoeDsaIndexer is unrelated to remote-code meta-initialization incompatibility." }, { - "left": "issue:43986", - "right": "issue:46726", + "left": "issue:44360", + "right": "issue:46762", "accept": false, - "reason": "Both involve processor/backend loading, but one is video AutoProcessor without torchvision and the other is unrecognized image processor handling." + "reason": "A missing ReLU in the DSA indexer and Minimax attention numerical mismatch are different bugs." }, { - "left": "issue:45381", - "right": "issue:46305", + "left": "issue:41628", + "right": "issue:43824", "accept": false, - "reason": "Different models and mechanisms: Qwen2.5-VL vision_position_ids vs MoonshineStreaming window masking." + "reason": "Both are import errors, but for different symbols and model paths; not the same underlying missing export." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "Code bug in DSA indexer vs a docstring issue for position_ids." + "reason": "One is a runtime indexing bug, the other is a docstring problem." }, { - "left": "issue:44279", - "right": "issue:46328", + "left": "issue:46305", + "right": "issue:46655", "accept": false, - "reason": "Generic dependency issue is unrelated to a SequenceFeatureExtractor.pad performance optimization." + "reason": "Sliding-window off-by-one and video padding calculation are separate processor bugs." }, { - "left": "issue:46305", - "right": "issue:46655", + "left": "issue:22355", + "right": "issue:28218", "accept": false, - "reason": "Both are video-processing bugs, but one is a sliding-window off-by-one and the other is padding calculation in a different model." + "reason": "Missing transformers.onnx module import and tokenizer spacing behavior are unrelated." }, { "left": "issue:22355", "right": "issue:27957", "accept": false, - "reason": "Missing transformers.onnx module import and Flash Attention 2 support for XLMRoberta are unrelated." + "reason": "Missing transformers.onnx module and XLMRoberta Flash Attention 2 are unrelated issues." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Import NameError in sentence-transformers does not share a code path with removed Sam3LiteText auto-mapping references." - }, - { - "left": "issue:41628", - "right": "issue:43824", - "accept": false, - "reason": "Different missing exports: AutoImageProcessor vs Qwen2_5_VLForConditionalGeneration." - }, - { - "left": "issue:22355", - "right": "issue:28218", - "accept": false, - "reason": "Import failure for transformers.onnx is unrelated to tokenizer spacing after added tokens." + "reason": "Import-time NameError for nn and stale auto_mappings references are different breakages." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "Server request handling crash and missing _tokenizer on Gemma4Processor are different failures." + "reason": "A chat-completions API rejection and a missing _tokenizer attribute in Gemma4Processor are different failures." }, { - "left": "issue:42831", - "right": "issue:45310", + "left": "issue:43756", + "right": "issue:44060", "accept": false, - "reason": "FineGrainedFP8 accuracy regression and Qwen3.5 MoE from_pretrained error are different problem types." + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are separate model-architecture bugs." }, { - "left": "issue:43576", - "right": "issue:43784", + "left": "issue:42831", + "right": "issue:45310", "accept": false, - "reason": "Broken transformers env CLI and a missing nn import in sentence-transformers are unrelated." + "reason": "FineGrainedFP8 accuracy drift and Qwen3.5 Moe from_pretrained error are not the same bug." }, { "left": "issue:44964", "right": "issue:47059", "accept": false, - "reason": "Phi-4 multimodal model loading and a huggingface_hub CLI test failure are not the same bug." + "reason": "Phi-4 multimodal loading and a CLI incompatibility with huggingface_hub are unrelated." }, { - "left": "issue:43756", - "right": "issue:44060", + "left": "issue:43576", + "right": "issue:43784", "accept": false, - "reason": "Smollm3 RoPE-layer issue and Qwen3-Next tied-weights warning are different model-specific regressions." + "reason": "Broken transformers env CLI and a missing nn import are different surfaces and causes." }, { - "left": "issue:44261", - "right": "issue:45092", + "left": "issue:44188", + "right": "issue:44373", "accept": false, - "reason": "Missing rms_norm_eps in MLA layernorm is unrelated to InternVL2 meta initialization." + "reason": "torch.compile attention-kernel divergence and a wrong docstring are unrelated." }, { - "left": "issue:44263", - "right": "issue:44360", + "left": "issue:44261", + "right": "issue:45092", "accept": false, - "reason": "Same subsystem name, but torch.split return values and missing ReLU are distinct defects." + "reason": "Missing rms_norm_eps in MLA q_a_layernorm and old InternVL2 meta-init incompatibility are different issues." }, { - "left": "issue:44188", - "right": "issue:44373", + "left": "issue:42796", + "right": "issue:43122", "accept": false, - "reason": "torch.compile attention divergence is unrelated to a wrong docstring." + "reason": "Both involve tokenization regressions, but one is model-specific BPE-marker handling while the other is a broader version-to-version behavior change." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "Different Qwen3 bugs: FineGrainedFP8 accuracy vs Qwen3-VL-Moe from_pretrained error." + "reason": "FineGrainedFP8 accuracy issues and Qwen3-VL-Moe loading/behavior problems are not the same underlying defect." }, { - "left": "issue:44829", - "right": "issue:46002", + "left": "issue:44263", + "right": "issue:44360", "accept": false, - "reason": "Training degradation under flash_attention_3 and NaNs after checkout+init are not the same code-path issue." + "reason": "Two different DSA-indexer bugs: one about torch.split outputs, one about a missing ReLU." }, { - "left": "issue:43986", - "right": "issue:47217", + "left": "issue:43644", + "right": "issue:46305", "accept": false, - "reason": "AutoProcessor backend-missing crash differs from VL failure when the image placeholder is in user text." + "reason": "Non-persistent buffer junk filling and a sliding-window mask off-by-one are separate regressions." }, { - "left": "issue:42796", - "right": "issue:43122", + "left": "issue:43986", + "right": "issue:47217", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different behaviors and likely different root causes." + "reason": "A torchvision-missing crash in AutoProcessor and VL failures when the placeholder is in user text have different triggers." }, { - "left": "issue:43644", - "right": "issue:46305", + "left": "issue:44829", + "right": "issue:46002", "accept": false, - "reason": "Non-persistent buffer initialization junk and sliding-window off-by-one are unrelated bugs." + "reason": "FlashAttention-3 degenerate training and NaNs after checkout+init are different numerical/training problems." }, { - "left": "issue:42831", - "right": "issue:45753", + "left": "issue:43540", + "right": "issue:46909", "accept": false, - "reason": "FP8 accuracy regression is unrelated to Qwen3_5 infinite-loop behavior on a specific image." + "reason": "Qwen3OmniMoe video-input ValueError and a broken DIA documentation example are unrelated." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Wrong offset_mapping and missing BOS/EOS insertion are different tokenizer issues." + "reason": "Wrong offset_mapping and missing BOS/EOS with add_special_tokens are distinct tokenizer bugs." }, { - "left": "issue:43540", - "right": "issue:46909", + "left": "issue:42831", + "right": "issue:45753", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError and a broken DIA documentation example are unrelated." + "reason": "FP8 accuracy drift and Qwen3_5 infinite looping on an image are different failure modes." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "Model-type warning for sam3_tracker and memory overflow in sam3-litetext are different issues." + "reason": "sam3_video vs sam3_tracker warning and sam3-litetext memory overflow are different sam3-family issues." }, { "left": "issue:45381", "right": "issue:46909", "accept": false, - "reason": "Qwen2.5-VL vision_position_ids bug is unrelated to a broken documentation example." + "reason": "Incorrect vision_position_ids in Qwen2.5-VL and a DIA docs bug are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong tokenizer decoder type are both v5 tokenizer issues, but not the same code-path problem." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Tokenizer offline-cache loading and quantized model saving are unrelated." } ] }, @@ -16774,6 +16865,7 @@ "nodes": [ "issue:22355", "issue:27957", + "issue:28348", "issue:29159", "issue:30990", "issue:33357", @@ -16782,10 +16874,7 @@ "issue:41628", "issue:42371", "issue:42757", - "issue:42886", "issue:42907", - "issue:42994", - "issue:43066", "issue:43116", "issue:43122", "issue:43329", @@ -16793,6 +16882,7 @@ "issue:43452", "issue:43475", "issue:43646", + "issue:43701", "issue:43824", "issue:43906", "issue:43976", @@ -16802,8 +16892,6 @@ "issue:44295", "issue:44297", "issue:44351", - "issue:44360", - "issue:44485", "issue:44488", "issue:44521", "issue:44625", @@ -16811,6 +16899,7 @@ "issue:44829", "issue:44945", "issue:44964", + "issue:44995", "issue:45200", "issue:45216", "issue:45278", @@ -16819,255 +16908,250 @@ "issue:46002", "issue:46172", "issue:46326", + "issue:46640", + "issue:46697", "issue:47059" ], "soft_pairs": [ - "issue:42886|issue:43066", - "issue:42886|issue:42994", - "issue:27957|issue:30990", - "issue:42757|issue:45278", "issue:43646|issue:46002", - "issue:30990|issue:33357", "issue:43329|issue:47059", "issue:43122|issue:46326", - "issue:39379|issue:46172", + "issue:27957|issue:30990", + "issue:30990|issue:33357", + "issue:42757|issue:45278", "issue:44625|issue:44829", + "issue:39379|issue:46172", "issue:22355|issue:29159", - "issue:42907|issue:44964", "issue:39401|issue:44521", - "issue:44295|issue:44297", - "issue:44117|issue:44704", + "issue:42907|issue:44964", "issue:39401|issue:44488", + "issue:44117|issue:44704", + "issue:44295|issue:44297", "issue:41628|issue:45278", + "issue:43122|issue:44521", "issue:42907|issue:44945", "issue:43976|issue:44945", - "issue:43122|issue:44521", "issue:44704|issue:45290", - "issue:43408|issue:43475", - "issue:42371|issue:44263", "issue:42907|issue:43452", + "issue:43408|issue:43475", "issue:43824|issue:44351", - "issue:44038|issue:44297", + "issue:42371|issue:44263", "issue:43116|issue:45200", - "issue:45216|issue:45478", - "issue:44038|issue:44295", + "issue:44038|issue:44297", "issue:43122|issue:43906", - "issue:44360|issue:44485" + "issue:45216|issue:45478", + "issue:28348|issue:46640", + "issue:44995|issue:46697", + "issue:43701|issue:44295" ], - "prepared_review_unit_hash": "25d4c17cc1966440e70e7c439af28b94fd16dbf2e2be1d17f0d189a40b95f473", + "prepared_review_unit_hash": "edb5ed92be789fac0841be4b3379d47fd33059ac8c60b8d475ca96ead22990c8", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36563, - "estimated_input_tokens": 9141, - "estimated_eval_tokens": 18538 + "soft_pair_count": 30, + "serialized_chars": 36297, + "estimated_input_tokens": 9075, + "estimated_eval_tokens": 18406 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is highly heterogeneous: most soft-similarity pairs share only broad keywords (tokenizer, import, save/load, multimodal) but not the same concrete bug or change. I would not merge any of the soft pairs. The strongest issue anchor is the active multimodal chat-template bug, which has the clearest discussion and inbound references.", - "confidence": 0.91, - "canonical_issue_reason": "issue:44521 is the best anchor: it is open, has the most inbound references/activity in this set, and describes a specific, reproducible multimodal chat-template failure rather than a broad symptom.", + "summary": "This cluster is mostly heterogeneous. The soft pairs share broad vocabulary (tokenizer, import errors, flash attention, multimodal models), but they do not describe the same underlying bug/change closely enough to merge as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45278 is the broadest user-facing regression report here (multiple import errors after upgrading), so it is the closest thing to a cluster anchor, though the cluster itself is not a true duplicate set.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 is the most suitable representative overall because it is the most active and specific unresolved bug in the group.", + "best_issue_reason": "issue:45278 has the clearest broad regression framing and is the most likely to represent the common failure mode among the mixed import/version-related reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:42886", - "right": "issue:43066", - "accept": false, - "reason": "Both involve tokenizers, but one is an offline cache-loading failure and the other is a decoder-type mismatch in v5; different code paths." - }, - { - "left": "issue:42886", - "right": "issue:42994", + "left": "issue:43646", + "right": "issue:46002", "accept": false, - "reason": "Offline tokenizer cache loading and quantized-model saving are unrelated bugs." + "reason": "Different failures: custom model init breakage vs NaNs in classification-head init/checkout. Same era, not the same code-path bug." }, { - "left": "issue:27957", - "right": "issue:30990", + "left": "issue:43329", + "right": "issue:47059", "accept": false, - "reason": "Flash Attention 2 for XLM-Roberta and Sentence Transformers hanging on load are different failures." + "reason": "Unrelated subsystems: multimodal token counting bug vs CLI completion attribute breakage from huggingface_hub." }, { - "left": "issue:42757", - "right": "issue:45278", + "left": "issue:43122", + "right": "issue:46326", "accept": false, - "reason": "Missing one dependency symbol is not the same as a broad set of import errors after an upgrade." + "reason": "Both tokenizer-related, but one is version-to-version tokenization drift and the other is continue_final_message behavior for mistral_common. Not the same defect." }, { - "left": "issue:43646", - "right": "issue:46002", + "left": "issue:27957", + "right": "issue:30990", "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are different problems." + "reason": "Flash Attention 2 support request for XLM-Roberta vs a loading hang in Sentence Transformers. Different models and symptoms." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Stuck loading versus bus error on macOS are different symptoms and likely different causes." + "reason": "Loading hang vs Mac bus error on a CLIP community model. No shared concrete bug path." }, { - "left": "issue:43329", - "right": "issue:47059", + "left": "issue:42757", + "right": "issue:45278", "accept": false, - "reason": "Undefined multimodal-token helper variables and a CLI compatibility error with huggingface_hub are unrelated." + "reason": "Both mention imports, but one is a specific missing symbol from huggingface_hub and the other is a broad upgrade regression. Not the same issue." }, { - "left": "issue:43122", - "right": "issue:46326", + "left": "issue:44625", + "right": "issue:44829", "accept": false, - "reason": "Tokenizer-version output differences and a mistral_common continue_final_message bug are not the same concrete issue." + "reason": "Qwen3.5 label propagation bug vs flash_attention_3 training degeneration. Different model behavior and code paths." }, { "left": "issue:39379", "right": "issue:46172", "accept": false, - "reason": "An mllama mask-handling bug and a DeepSeek-V4 tensor operation issue are different code paths." + "reason": "Separate multimodal/math bugs in different models; no shared failing function or fix." }, { - "left": "issue:44625", - "right": "issue:44829", + "left": "issue:22355", + "right": "issue:29159", "accept": false, - "reason": "Qwen3.5 num_labels propagation and flash_attention_3 training degeneration are unrelated." + "reason": "Missing transformers.onnx module import vs slow/fast tokenizer inconsistency. Different components and failure modes." }, { - "left": "issue:22355", - "right": "issue:29159", + "left": "issue:39401", + "right": "issue:44521", "accept": false, - "reason": "Missing transformers.onnx module and slow/fast tokenizer inconsistency are not the same bug." + "reason": "Both concern tokenization, but one is offset_mapping for Qwen3 and the other is assistant masks in multimodal chat templating." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Quantized-model save failures and a multimodal model load failure are different issues." + "reason": "Saving dequantized checkpoints vs loading Phi-4 multimodal. Different model families and unrelated code paths." }, { "left": "issue:39401", - "right": "issue:44521", - "accept": false, - "reason": "Wrong offset_mapping and all-zero assistant masks are both tokenizer-related but affect different APIs and behavior." - }, - { - "left": "issue:44295", - "right": "issue:44297", + "right": "issue:44488", "accept": false, - "reason": "Reading a buffer-registered position_ids and tokenizer_class mismatch on save_pretrained are unrelated." + "reason": "Wrong offset mapping in Qwen3 tokenizer vs failure to load a specific BERT model. Not the same defect." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "A tokenizer mapping returning None and AutoProcessor not forwarding kwargs to cached_file are different call sites." + "reason": "Tokenizer mapping fallback bug vs AutoProcessor kwargs propagation. Different APIs and root causes." }, { - "left": "issue:39401", - "right": "issue:44488", + "left": "issue:44295", + "right": "issue:44297", "accept": false, - "reason": "Wrong offset mappings and a model load failure for cjvt/sleng-bert are different bugs." + "reason": "Reading a buffer-registered position_ids error vs tokenizer_class mismatch on save_pretrained. Different serialization/state issues." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "A specific AutoImageProcessor import error is not the same as a broad version-upgrade import regression." + "reason": "Import name error for AutoImageProcessor vs a broad post-upgrade import breakage. Related only at a high level." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer drift across versions vs multimodal assistant mask generation. Same broad area, different bugs." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Saving dequantized models and incorrect pipeline-parallel LLM output are unrelated." + "reason": "Checkpoint save/dequantization failure vs incorrect pipeline-parallel inference output. No shared concrete fix." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python-version compatibility problems and pipeline-parallel output corruption are different issues." + "reason": "Python-version compatibility regression vs pipeline-parallel output corruption. Different symptoms and code paths." }, { - "left": "issue:43122", - "right": "issue:44521", + "left": "issue:44704", + "right": "issue:45290", "accept": false, - "reason": "Tokenizer output drift across versions and multimodal assistant-mask generation are distinct behaviors." + "reason": "Processor metadata/download kwargs bug vs chat_template crash on tool-call messages. Distinct functionality." }, { - "left": "issue:44704", - "right": "issue:45290", + "left": "issue:42907", + "right": "issue:43452", "accept": false, - "reason": "Not forwarding kwargs to cached_file and crashes on assistant tool-call messages are different chat-template/processor bugs." + "reason": "Dequantized save path vs gguf/from_pretrained loading breakage. Different model-loading workflows." }, { "left": "issue:43408", "right": "issue:43475", "accept": false, - "reason": "A model-type warning and a missing attribute on Sam3VisionEncoderOutput are different SAM 3 video issues." + "reason": "SAM3 model-type warning vs missing attribute in SAM3 video encoder output. Related model family, but not the same bug." }, { - "left": "issue:42371", - "right": "issue:44263", + "left": "issue:43824", + "right": "issue:44351", "accept": false, - "reason": "TF32 API guidance and a torch.split return-value issue are unrelated." + "reason": "Different missing imports: Qwen2.5-VL class vs HybridCache. Same broad import category, not one duplicate." }, { - "left": "issue:42907", - "right": "issue:43452", + "left": "issue:42371", + "right": "issue:44263", "accept": false, - "reason": "Quantized save failures and gguf_file breaking from_pretrained are different loading/saving paths." + "reason": "TF32 API warning vs a torch.split return-value issue in GlmMoeDsaIndexer. Unrelated." }, { - "left": "issue:43824", - "right": "issue:44351", + "left": "issue:43116", + "right": "issue:45200", "accept": false, - "reason": "Missing Qwen2_5_VLForConditionalGeneration and missing HybridCache are separate import regressions, not one concrete bug." + "reason": "Example script multi-label output bug vs Gemma 4 mm_token_type_ids defaulting. Different failure points." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "Qwen3-VL-Moe compatibility and tokenizer_class mismatch on save_pretrained are unrelated." + "reason": "Qwen3-VL-Moe compatibility issue vs tokenizer save_pretrained metadata mismatch. Different problems." }, { - "left": "issue:43116", - "right": "issue:45200", + "left": "issue:43122", + "right": "issue:43906", "accept": false, - "reason": "Empty multi-label classification outputs and mm_token_type_ids defaulting are different example/model issues." + "reason": "Tokenizer behavior change across versions vs isolated reproduction of another issue. No concrete overlap." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "A bad saved checkpoint and a from_pretrained error for Qwen3.5 are related by model family but not the same concrete failure." + "reason": "Both involve Qwen3.5 and version 5.4.0, but one is a bad saved checkpoint and the other is a from_pretrained loading error. Related area, not clearly the same bug." }, { - "left": "issue:44038", - "right": "issue:44295", + "left": "issue:28348", + "right": "issue:46640", "accept": false, - "reason": "Qwen3-VL-Moe support and buffer registration of position_ids are unrelated." + "reason": "Specific GPT2 Flash Attention 2 support request vs a generic flash-attention issue report. Not the same code path." }, { - "left": "issue:43122", - "right": "issue:43906", + "left": "issue:44995", + "right": "issue:46697", "accept": false, - "reason": "Version-to-version tokenization differences and an isolated reproduction of another issue are not the same bug." + "reason": "GlmMoeDsa stale cache on second forward pass vs Swin2SR image padding overflow. Completely different models." }, { - "left": "issue:44360", - "right": "issue:44485", + "left": "issue:43701", + "right": "issue:44295", "accept": false, - "reason": "A missing ReLU in DSA indexing and GLM-5 RoPE implementation are different model internals." + "reason": "resume_from_checkpoint key mismatch vs position_ids buffer access error. Different state-loading bugs." } ] }, @@ -17092,7 +17176,6 @@ "cluster_id": "cluster-10000-544", "nodes": [ "issue:22355", - "issue:28348", "issue:29127", "issue:36296", "issue:38617", @@ -17110,15 +17193,18 @@ "issue:43611", "issue:43976", "issue:44038", + "issue:44075", "issue:44188", "issue:44222", "issue:44351", + "issue:44360", "issue:44483", + "issue:44485", "issue:44610", "issue:44625", + "issue:44936", "issue:44938", "issue:44945", - "issue:44995", "issue:45072", "issue:45083", "issue:45278", @@ -17131,8 +17217,6 @@ "issue:46231", "issue:46328", "issue:46614", - "issue:46640", - "issue:46697", "issue:46762", "issue:46877", "issue:46909", @@ -17140,210 +17224,217 @@ "issue:47312" ], "soft_pairs": [ - "issue:44995|issue:46697", - "issue:44038|issue:44222", "issue:43278|issue:44945", - "issue:46172|issue:46762", - "issue:42175|issue:44938", "issue:22355|issue:29127", - "issue:28348|issue:46640", - "issue:42222|issue:47312", + "issue:44038|issue:44222", + "issue:44360|issue:44485", + "issue:42175|issue:44938", + "issue:46172|issue:46762", "issue:43493|issue:45753", - "issue:39401|issue:46614", + "issue:42222|issue:47312", "issue:39379|issue:43232", + "issue:39401|issue:46614", + "issue:42673|issue:45072", "issue:44188|issue:46328", - "issue:42757|issue:44351", "issue:45357|issue:45478", - "issue:42673|issue:45072", + "issue:42757|issue:44351", "issue:36296|issue:43116", "issue:46231|issue:46877", "issue:44483|issue:47217", - "issue:42898|issue:46614", "issue:44625|issue:46002", "issue:44610|issue:46909", - "issue:43976|issue:45446", + "issue:42898|issue:46614", "issue:45083|issue:45446", + "issue:43976|issue:45446", + "issue:43611|issue:46002", "issue:38617|issue:45278", - "issue:43611|issue:46002" + "issue:43116|issue:43493", + "issue:44075|issue:44936" ], - "prepared_review_unit_hash": "156fb9f2d3566b634993d3616c47e6f9bd0d19f3be6943fae394c40dd1edb3d7", + "prepared_review_unit_hash": "2f9dabe17227011267e7dec9597bb262e53193e3a6366c2413a030ae018cba09", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 25, - "serialized_chars": 34086, - "estimated_input_tokens": 8522, - "estimated_eval_tokens": 17300 + "soft_pair_count": 26, + "serialized_chars": 34327, + "estimated_input_tokens": 8582, + "estimated_eval_tokens": 17420 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft edge looks like a true duplicate: the proposed pairs mostly share only a broad subsystem name or symptom class (imports, tokenizers, attention, Qwen3, etc.) but point to different concrete bugs or feature requests. I would not merge any of these issues as duplicates.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This set is mostly a grab bag of unrelated Hugging Face Transformers issues. Several pairs share a theme or model family, but none look like the same concrete bug/change closely enough to deduplicate, so no soft edge should be accepted.", + "confidence": 0.91, + "canonical_issue_reason": "issue:39401 is a clean, specific bug report with a narrowly defined symptom (wrong offset_mapping) and clear reproduction context, making it a strong representative issue among the set.", "canonical_pr_reason": null, - "best_issue_reason": "No single canonical duplicate emerges; the issues cover distinct failures across unrelated code paths. If forced to pick the broadest report, issue 45278 is the most umbrella-like, but it is not a clear duplicate target.", + "best_issue_reason": "issue:39401 is the strongest standalone issue: concrete, model-specific, and directly actionable without being overly broad or speculative.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44995", - "right": "issue:46697", + "left": "issue:43278", + "right": "issue:44945", "accept": false, - "reason": "Different models and failures: stale cache crash on second forward pass vs image processor over-padding." + "reason": "Both involve training/evaluation behavior, but one is a dtype flip and the other is pipeline-parallel output correctness; different code paths and symptoms." }, { - "left": "issue:44038", - "right": "issue:44222", + "left": "issue:22355", + "right": "issue:29127", "accept": false, - "reason": "Both involve MoE/FP8 wording, but one is a Qwen3-VL-Moe compatibility issue and the other is save_pretrained behavior; not the same bug." + "reason": "Completely different failures: missing transformers.onnx module vs a LayoutLMv3 error-message clarity problem." }, { - "left": "issue:43278", - "right": "issue:44945", + "left": "issue:44038", + "right": "issue:44222", "accept": false, - "reason": "Embedding dtype mismatch during eval is unrelated to pipeline-parallel output corruption." + "reason": "Both mention MoE-related regressions, but one is a Qwen3-VL-Moe loading bug and the other is FP8 save_pretrained behavior; not the same concrete bug." }, { - "left": "issue:46172", - "right": "issue:46762", + "left": "issue:44360", + "right": "issue:44485", "accept": false, - "reason": "Both mention numerical/transpose behavior in attention, but they are different models and code paths." + "reason": "Unrelated technical areas: DSA indexer ReLU discussion vs GLM-5 RoPE implementation." }, { "left": "issue:42175", "right": "issue:44938", "accept": false, - "reason": "Package installation backend omission is unrelated to Python 3.14 loading failure." + "reason": "Both are import/platform issues, but one is an extras/backend dependency omission and the other is a Python 3.14 load failure; different root causes." }, { - "left": "issue:22355", - "right": "issue:29127", + "left": "issue:46172", + "right": "issue:46762", "accept": false, - "reason": "Importing transformers.onnx and LayoutLMv3 box error messaging are unrelated." + "reason": "Both concern numerical behavior, but they affect different models and mechanisms; no evidence of the same underlying bug." }, { - "left": "issue:28348", - "right": "issue:46640", + "left": "issue:43493", + "right": "issue:45753", "accept": false, - "reason": "A GPT-2 flash-attention feature request is not the same as a generic Flash Attention 2 bug." + "reason": "Different model families and symptoms: SigLIP2 implementation discrepancy vs Qwen3_5 infinite-loop generation on a specific image." }, { "left": "issue:42222", "right": "issue:47312", "accept": false, - "reason": "VitPose model breakage is unrelated to kernel 0.16 incompatibility." + "reason": "VitPose model breakage and a kernels-version compatibility issue are unrelated." }, { - "left": "issue:43493", - "right": "issue:45753", + "left": "issue:39379", + "right": "issue:43232", "accept": false, - "reason": "SigLIP2 implementation mismatch and Qwen3_5 infinite-loop-on-image are different bugs." + "reason": "One is an mllama masking bug, the other is generation kwargs after sync_gpus; different code paths." }, { "left": "issue:39401", "right": "issue:46614", "accept": false, - "reason": "Tokenizer offset mapping and stale phonemizer backend are different tokenizer bugs." + "reason": "Tokenizer offset mapping and Wav2Vec2 phonemizer backend caching are distinct tokenizer/phonemizer bugs." }, { - "left": "issue:39379", - "right": "issue:43232", + "left": "issue:42673", + "right": "issue:45072", "accept": false, - "reason": "mllama mask handling is unrelated to generation kwargs after sync_gpus." + "reason": "VRAM leakage in multi-threaded Qwen3ForCausalLM use is unrelated to dtype mismatches in inference." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "Attention-kernel divergence on torch.compile is unrelated to a padding performance issue." - }, - { - "left": "issue:42757", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but for different missing symbols in different packages." + "reason": "Attention-kernel branching under torch.compile is unrelated to SequenceFeatureExtractor.pad performance overhead." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Both are Qwen3.5-related, but save_pretrained key regression and from_pretrained error are distinct failures." + "reason": "Both involve Qwen3.5, but one is incorrect save_pretrained keys and the other is a from_pretrained import/load error in newer Transformers." }, { - "left": "issue:42673", - "right": "issue:45072", + "left": "issue:42757", + "right": "issue:44351", "accept": false, - "reason": "VRAM leak in dataloader threads is unrelated to dtype mismatches in inference." + "reason": "Both are import errors, but for different symbols from different packages; no shared underlying bug is established." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training bug is unrelated to the multi-label example script returning empty results." + "reason": "Tensor-parallel training bug and multi-label classification example returning empty results are unrelated." }, { "left": "issue:46231", "right": "issue:46877", "accept": false, - "reason": "Documentation parameter-name mismatch is not the same as doc-lint output at import time." + "reason": "Documentation parameter mismatch is not the same as auto_docstring printing diagnostics at import time." }, { "left": "issue:44483", "right": "issue:47217", "accept": false, - "reason": "Chat completions API rejection and image-placeholder handling in VL models are different request-processing bugs." - }, - { - "left": "issue:42898", - "right": "issue:46614", - "accept": false, - "reason": "Tokenization cleanup behavior change and phonemizer backend state are unrelated." + "reason": "The chat/completions request rejection issue and the VL image-placeholder failure are different API/model behaviors." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "num_labels propagation bug is unrelated to NaNs in classification heads." + "reason": "Config propagation for Qwen3.5 num_labels and NaNs in classification heads are different bugs with different failure modes." }, { "left": "issue:44610", "right": "issue:46909", "accept": false, - "reason": "Wrong processor output size and broken DIA documentation example are unrelated." + "reason": "Processor input-size mismatch and a broken DIA docs example are not the same issue." }, { - "left": "issue:43976", - "right": "issue:45446", + "left": "issue:42898", + "right": "issue:46614", "accept": false, - "reason": "Python version support breakage is unrelated to an incorrect PyTorch version check for AuxRequest." + "reason": "Tokenization cleanup semantics and phonemizer backend state retention are unrelated tokenizer/audio bugs." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Qwen3_omni_moe feature-length helper behavior is unrelated to flex_attention import gating." + "reason": "A feature-length calculation bug in qwen3_omni_moe is unrelated to a PyTorch version-check issue in flex_attention." }, { - "left": "issue:38617", - "right": "issue:45278", + "left": "issue:43976", + "right": "issue:45446", "accept": false, - "reason": "A specific missing import is not the same as a broad 'many import errors' regression, even if they may share a release." + "reason": "Python-version compatibility and PyTorch version gating are separate environment-compatibility problems." }, { "left": "issue:43611", "right": "issue:46002", "accept": false, - "reason": "Base_model_prefix loading regression is unrelated to NaNs after checkout/init in classification heads." + "reason": "base_model_prefix loading regression and NaNs in classification heads are different failures." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing import name and a broad set of import errors after upgrade may be related at a high level, but not enough to treat as the same concrete bug." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script classification output bug and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling and trainer.evaluate() failing after train() are separate training-loop issues." } ] }, @@ -17378,27 +17469,27 @@ "issue:43066", "issue:43116", "issue:43208", - "issue:43493", + "issue:43408", "issue:43618", "issue:43673", "issue:43688", + "issue:43742", "issue:43824", "issue:43825", "issue:43827", "issue:43874", "issue:43937", "issue:43992", - "issue:44075", "issue:44117", "issue:44291", "issue:44355", + "issue:44466", "issue:44485", "issue:44704", "issue:44779", - "issue:44936", + "issue:44977", "issue:45071", "issue:45081", - "issue:45092", "issue:45341", "issue:45362", "issue:45405", @@ -17406,11 +17497,11 @@ "issue:45478", "issue:45563", "issue:45588", + "issue:46002", "issue:46231", "issue:46328", "issue:46710", "issue:46726", - "issue:46762", "issue:46829", "issue:46894", "issue:46981", @@ -17418,230 +17509,230 @@ ], "soft_pairs": [ "issue:43066|issue:44779", - "issue:43116|issue:43493", "issue:45362|issue:45478", - "issue:44075|issue:44936", - "issue:43937|issue:44485", "issue:42673|issue:42796", + "issue:43937|issue:44485", "issue:43066|issue:46710", - "issue:44355|issue:46231", "issue:46726|issue:46981", "issue:43618|issue:46829", "issue:43116|issue:43825", - "issue:42371|issue:43874", + "issue:44355|issue:46231", "issue:43827|issue:45446", + "issue:42371|issue:43874", "issue:43992|issue:44704", "issue:44291|issue:45071", - "issue:45405|issue:45446", - "issue:43824|issue:46981", "issue:44117|issue:45588", + "issue:45405|issue:45446", "issue:39379|issue:43688", - "issue:46328|issue:47116", + "issue:43824|issue:46981", "issue:44117|issue:45081", + "issue:42617|issue:45341", "issue:43673|issue:46328", + "issue:46328|issue:47116", "issue:33453|issue:39401", "issue:43064|issue:43673", "issue:43208|issue:46894", "issue:43937|issue:45563", - "issue:42617|issue:45341", - "issue:45092|issue:46762" + "issue:43408|issue:43742", + "issue:44466|issue:46002", + "issue:43937|issue:44977" ], - "prepared_review_unit_hash": "2182b947be908dfb498d6bd41532315244408696675f86986652c95febef816c", + "prepared_review_unit_hash": "4c73b4a067e89f1550684077b77f406968ef30c59d0966f40df259899d8b6c42", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35948, - "estimated_input_tokens": 8987, - "estimated_eval_tokens": 18230 + "serialized_chars": 36003, + "estimated_input_tokens": 9001, + "estimated_eval_tokens": 18258 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of unrelated issues, mostly around Transformers v5 tokenizer/model-loading regressions. There is no true duplicate core; the closest thematic group is tokenizer-loading bugs, but the concrete failures differ across models and code paths.", - "confidence": 0.68, - "canonical_issue_reason": "Issue 43066 is the broadest and most central representative of the set: it is a direct Transformers v5 tokenizer-loading regression and sits near several other tokenizer-related reports, even though it does not truly subsume them.", + "summary": "This cluster is mostly a loose thematic group of unrelated bug reports, with only broad overlap around tokenizer/v5 regressions. None of the soft pairs look like the same concrete underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.42, + "canonical_issue_reason": "issue:43066 is the most generally useful representative because it is a concrete tokenizer-v5 regression rather than a highly model-specific symptom, and several nearby items are tokenizer-related but not identical.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43066 is the best single issue to keep as a representative umbrella for the cluster\u2019s strongest shared theme (v5 tokenizer/loading regressions).", + "best_issue_reason": "issue:43066 is the strongest standalone issue in the set: specific enough to be actionable, but broad enough to serve as the clearest representative of the tokenizer-regression subset.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer-related v5 regressions, but one is a decoder-type change and the other is incorrect DeepSeek tokenization output; different concrete bugs and code paths." - }, - { - "left": "issue:43116", - "right": "issue:43493", - "accept": false, - "reason": "One is an example-script multi-label classification bug; the other is a SigLIP2 implementation discrepancy. No shared underlying failure." + "reason": "Both are tokenizer-related v5 regressions, but one is about decoder type selection and the other about incorrect DeepSeek tokenization results; different failure modes and likely different fixes." }, { "left": "issue:45362", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3.5, but one is a chat crash and the other is a from_pretrained error for MoE weights; not the same bug." + "reason": "Same model family, but one is a chat crash and the other is a from_pretrained error; different code paths and symptoms." }, { - "left": "issue:44075", - "right": "issue:44936", + "left": "issue:42673", + "right": "issue:42796", "accept": false, - "reason": "SGD argument handling in optimizers and trainer.evaluate() failing after train are separate training-loop issues." + "reason": "Completely different bugs: VRAM leak in dataloader threads vs tokenizer BPE marker handling." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "GLM-5 GenerationConfig validation and GLM-5 RoPE implementation are distinct model issues with different failure modes." - }, - { - "left": "issue:42673", - "right": "issue:42796", - "accept": false, - "reason": "VRAM leak in multi-dataloader-thread usage and tokenizer BPE-marker handling are unrelated subsystems." + "reason": "GLM-5 appears in both, but generation config validation and RoPE implementation are unrelated code paths." }, { "left": "issue:43066", "right": "issue:46710", "accept": false, - "reason": "One is a tokenizer decoder-type regression; the other is an accuracy regression on DeepSeek after upgrade. Same version context, different bug." - }, - { - "left": "issue:44355", - "right": "issue:46231", - "accept": false, - "reason": "Compiled Python file errors and a docs parameter-name mismatch are not the same underlying issue." + "reason": "Both are v5/DeepSeek-adjacent regressions, but one is a tokenizer decoder-type issue and the other is an accuracy regression; not the same bug." }, { "left": "issue:46726", "right": "issue:46981", "accept": false, - "reason": "AutoProcessor vision-backend fallback and repeated HTTP 429 retries are both in AutoProcessor, but they affect different code paths and symptoms." + "reason": "AutoProcessor backend-missing handling is unrelated to HTTP 429 retry behavior." }, { "left": "issue:43618", "right": "issue:46829", "accept": false, - "reason": "Both are output-field assignment bugs, but they concern different model outputs and different missing/incorrect fields." + "reason": "Both concern model output fields, but CLIP attentions and VideoPrism hidden_states are distinct output-shape/assignment bugs." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "An empty-results example script bug and an incorrect pipeline error message are separate issues." + "reason": "One is a multi-label classification example bug; the other is an incorrect pipeline error message. Different subsystems and fixes." }, { - "left": "issue:42371", - "right": "issue:43874", + "left": "issue:44355", + "right": "issue:46231", "accept": false, - "reason": "TF32 control settings and a missing image-processor method are unrelated bugs." + "reason": "Running compiled Python files is unrelated to a documentation parameter-name mismatch." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing pipeline() and a PyTorch version check bug in flex_attention are unrelated." + "reason": "Docs still referencing pipeline() and a flex_attention PyTorch version check are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance and a GLM46V image-patch attribute error are unrelated." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "Both involve loading/pretrained flows, but one misses weights in UMT5Encoder and the other drops kwargs in AutoProcessor/cached_file." + "reason": "Missing encoder weights on load vs AutoProcessor kwargs propagation are different loader bugs." }, { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "init_empty_weights compatibility and PretrainedConfig type-checking breakage are different v5 API regressions." - }, - { - "left": "issue:45405", - "right": "issue:45446", - "accept": false, - "reason": "A dependency version bump not yet on PyPI and an incorrect PyTorch version guard are unrelated." + "reason": "Both are v5-related type/config issues, but they affect different APIs and failure mechanisms." }, { - "left": "issue:43824", - "right": "issue:46981", + "left": "issue:44117", + "right": "issue:45588", "accept": false, - "reason": "Qwen2.5-VL import failure and AutoProcessor 429 retry behavior are unrelated." + "reason": "Tokenizer mapping fallback and flash_attention s_aux handling are unrelated." }, { - "left": "issue:44117", - "right": "issue:45588", + "left": "issue:45405", + "right": "issue:45446", "accept": false, - "reason": "Tokenizer mapping None-handling and a flash_attention crash on s_aux=None are different bugs." + "reason": "A dependency-version bump problem is not the same as an incorrect PyTorch version gate." }, { "left": "issue:39379", "right": "issue:43688", "accept": false, - "reason": "Mllama masking logic and MoE auxiliary-loss normalization are separate numerical correctness issues." + "reason": "Different model families and different logic: masked-out mask handling vs auxiliary-loss normalization." }, { - "left": "issue:46328", - "right": "issue:47116", + "left": "issue:43824", + "right": "issue:46981", "accept": false, - "reason": "A pad() performance regression and a streaming pipeline regression are not the same code-path problem." + "reason": "ImportError for a missing model class and repeated 429 retries are unrelated." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Both touch tokenizer loading, but one is a mapping-None assumption and the other is a Mistral regex patch crash; different fixes." + "reason": "Both involve tokenizer loading, but one is mapping-name None handling and the other is a Mistral regex/backend_tokenizer crash; not the same fix." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A runtime failure in 3d_parallel.py is not the same as a small testing_utils bug." }, { "left": "issue:43673", "right": "issue:46328", "accept": false, - "reason": "GenerationMixin cache handling and SequenceFeatureExtractor.pad performance are unrelated." + "reason": "Generation cache handling and SequenceFeatureExtractor.pad performance are unrelated." + }, + { + "left": "issue:46328", + "right": "issue:47116", + "accept": false, + "reason": "Padding performance and streaming pipeline regression are different behaviors in different layers." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Both are tokenizer issues, but one is a generic loading regression and the other is a Qwen3 offset_mapping bug." + "reason": "Both are tokenizer-loading related, but one is a broad regression report and the other is a specific Qwen3 offset_mapping bug; not enough evidence they are the same underlying issue." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Wrong optimizer state placement in Trainer/FSDP2 and missing GenerationMixin cache in chunked_prefill are different failures." + "reason": "Training optimizer-state corruption on nonzero ranks is unrelated to chunked_prefill cache absence." }, { "left": "issue:43208", "right": "issue:46894", "accept": false, - "reason": "Both are training bugs, but they concern different models and different mechanics; not a single underlying defect." + "reason": "Both mention training bugs, but xLSTM's multiple model issues and CohereASR's double-shift loss bug are different concrete problems." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "GenerationConfig validation failure and a stale num_return_sequences warning are distinct generation-related issues." + "reason": "GenerationConfig validation and a stale warning for num_return_sequences are different issues." }, { - "left": "issue:42617", - "right": "issue:45341", + "left": "issue:43408", + "right": "issue:43742", "accept": false, - "reason": "Running 3d_parallel.py failing and a small testing_utils.py bug are unrelated." + "reason": "A model-type warning and a MobileLLM loading key error are unrelated." }, { - "left": "issue:45092", - "right": "issue:46762", + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of lm_head.weight depends on device; classification head NaNs after checkout/init are a different failure mode." + }, + { + "left": "issue:43937", + "right": "issue:44977", "accept": false, - "reason": "Meta-init incompatibility for old remote-code checkpoints and Minimax attention numerical differences are different bugs." + "reason": "Both touch GLM/Qwen-style generation, but one is a GenerationConfig validity error and the other is flash-attention generation failure; different code paths." } ] }, @@ -17670,41 +17761,42 @@ "issue:39401", "issue:41628", "issue:42222", + "issue:42371", + "issue:42831", + "issue:43097", "issue:43122", - "issue:43408", - "issue:43421", "issue:43441", "issue:43450", - "issue:43742", "issue:43937", "issue:43986", "issue:44220", "issue:44263", "issue:44380", "issue:44410", - "issue:44466", "issue:44483", + "issue:44493", "issue:44496", "issue:44704", "issue:44779", "issue:44829", "issue:44877", "issue:44964", - "issue:44977", "issue:45071", + "issue:45092", "issue:45216", "issue:45278", "issue:45310", "issue:45341", + "issue:45440", "issue:45464", "issue:45561", "issue:45632", - "issue:46002", "issue:46077", "issue:46231", "issue:46298", "issue:46328", "issue:46512", + "issue:46710", "issue:46752", "issue:46762", "issue:46897", @@ -17715,233 +17807,230 @@ ], "soft_pairs": [ "issue:42222|issue:47000", - "issue:44466|issue:46002", - "issue:44380|issue:46077", - "issue:43937|issue:44977", - "issue:43408|issue:43742", - "issue:42222|issue:43441", "issue:43450|issue:43986", + "issue:44380|issue:46077", "issue:45071|issue:46512", + "issue:45092|issue:46762", + "issue:42222|issue:43441", "issue:39379|issue:46897", "issue:42222|issue:46752", - "issue:44263|issue:46762", "issue:44877|issue:46231", + "issue:45632|issue:46981", "issue:39401|issue:44779", + "issue:44263|issue:46762", + "issue:44483|issue:45464", "issue:44829|issue:45216", - "issue:45632|issue:46981", "issue:45341|issue:45561", "issue:44410|issue:45310", "issue:45071|issue:47059", - "issue:44483|issue:45464", - "issue:45341|issue:46328", - "issue:39401|issue:43421", - "issue:44704|issue:47110", + "issue:45440|issue:46328", + "issue:39401|issue:46710", + "issue:38617|issue:44220", "issue:43937|issue:44496", "issue:43122|issue:46298", + "issue:44704|issue:47110", "issue:38617|issue:41628", - "issue:38617|issue:44220", - "issue:44964|issue:45278" + "issue:44964|issue:45278", + "issue:43097|issue:44493", + "issue:42371|issue:42831" ], - "prepared_review_unit_hash": "79f029f4f8e1c42109c805c13dd81e687cc4630b188c2003403d7e122d919e7a", + "prepared_review_unit_hash": "aa880f070434e00f8d8ac3f21c431b075c45685de315d3427c6393ed891814d0", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35044, - "estimated_input_tokens": 8761, - "estimated_eval_tokens": 17778 + "serialized_chars": 35562, + "estimated_input_tokens": 8891, + "estimated_eval_tokens": 18038 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most pairs are clearly unrelated across different models, APIs, or failure modes. The only accepted soft pair appears to describe the same Qwen3.5 chat/completions streaming bug, with issue:45464 giving the more specific description.", - "confidence": 0.86, - "canonical_issue_reason": "issue:45464 is the clearest representative because it explicitly names the Qwen3.5-0.8B chat/completions streaming failure; issue:44483 appears to be the same bug, but its visible title is less specific.", + "summary": "No clear duplicate cluster emerges here: the soft pairs are mostly broad-similarity matches across different models, APIs, or failure modes, so they should not be merged as the same underlying bug. I would treat these as unrelated issues with superficial overlap.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45464 is the best issue to keep as the representative report because it is the more concrete description of the shared chat/completions streaming failure.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42222", "right": "issue:47000", "accept": false, - "reason": "Different models and failure modes: VitPose breakage versus an RTDetr loss crash." + "reason": "Both mention model/runtime failures, but one is a VitPose import breakage and the other is an RTDetr loss crash; different code paths and symptoms." }, { - "left": "issue:44466", - "right": "issue:46002", + "left": "issue:43450", + "right": "issue:43986", "accept": false, - "reason": "Serialization of tied weights is unrelated to NaNs in classification heads after checkout/init." + "reason": "Both involve video processing, but one is an incorrect batched output shape and the other is an AutoProcessor crash without torchvision; different underlying bugs." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "Both involve attention behavior, but they are different bugs in different models and code paths." + "reason": "Both are attention-related, but one is GPT2 scaling ignored in SDPA/FlashAttention and the other is Gemma4 building the wrong mask type; not the same defect." }, { - "left": "issue:43937", - "right": "issue:44977", + "left": "issue:45071", + "right": "issue:46512", "accept": false, - "reason": "Invalid GenerationConfig is a different problem from Qwen3.5 flash-attention generation issues." + "reason": "One is a type-checking regression in v5.4.0, the other is dtype metadata staying stale after from_config; unrelated config/model initialization issues." }, { - "left": "issue:43408", - "right": "issue:43742", + "left": "issue:45092", + "right": "issue:46762", "accept": false, - "reason": "The sam3_video model-type warning and the MobileLLM key error are unrelated loading bugs." + "reason": "Both are model-behavior regressions, but one is meta-initialization incompatibility and the other is numerical mismatch in Minimax MSA attention; different failure mechanisms." }, { "left": "issue:42222", "right": "issue:43441", "accept": false, - "reason": "VitPose model breakage and Ministral-3 FlashAttention failure are distinct model-specific regressions." - }, - { - "left": "issue:43450", - "right": "issue:43986", - "accept": false, - "reason": "Incorrect batched video processor shapes and a missing torchvision crash are different video-related issues." - }, - { - "left": "issue:45071", - "right": "issue:46512", - "accept": false, - "reason": "PretrainedConfig type checking and stale config.dtype after from_config are separate config bugs." + "reason": "VitPose model breakage and Ministral FlashAttention failure are separate model-specific regressions with different stacks." }, { "left": "issue:39379", "right": "issue:46897", "accept": false, - "reason": "mllama mask handling and Florence2 training-loss double-shift are different model-specific bugs." + "reason": "Both are loss/masking-related, but one is mllama mask handling and the other is Florence2 training-loss double shift; not the same bug." }, { "left": "issue:42222", "right": "issue:46752", "accept": false, - "reason": "VitPose breakage and an empty-list apply_chat_template IndexError are unrelated." + "reason": "VitPose import/model breakage is unrelated to apply_chat_template crashing on an empty list." }, { - "left": "issue:44263", - "right": "issue:46762", + "left": "issue:44877", + "right": "issue:46231", "accept": false, - "reason": "torch.split return values in one model indexer and Minimax attention numerical drift are not the same bug." + "reason": "One is strict config loading for granite_speech, the other is a documentation parameter-name mismatch; no shared code-path bug." }, { - "left": "issue:44877", - "right": "issue:46231", + "left": "issue:45632", + "right": "issue:46981", "accept": false, - "reason": "Strict config loading and a documentation parameter-name typo are unrelated." + "reason": "Both involve loading/caching, but one is a local trust_remote_code cache collision and the other is repeated 429 retries in AutoProcessor; different causes." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer complaints, but they target different models and failure mechanisms." + "reason": "Both are tokenizer issues, but one is wrong offset_mapping and the other is different tokenization output in v5; same area, different bug reports." }, { - "left": "issue:44829", - "right": "issue:45216", + "left": "issue:44263", + "right": "issue:46762", "accept": false, - "reason": "FlashAttention training degeneration and save_pretrained checkpoint regression are different issues." + "reason": "A torch.split indexing/value issue is not the same as a Minimax attention numerical discrepancy." }, { - "left": "issue:45632", - "right": "issue:46981", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "Cache-path collision for local models and repeated 429 retries are unrelated AutoProcessor behaviors." + "reason": "One is chat/completions request validation in v5.3, the other is streaming inference failure for Qwen3.5-0.8B; different API layers and failure modes." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "One is a flash_attention_3 training degeneration issue, the other is incorrect save_pretrained output for Qwen3.5; different behaviors and code paths." }, { "left": "issue:45341", "right": "issue:45561", "accept": false, - "reason": "A testing_utils bug and an xdist race on captured_info.txt are not the same underlying issue." + "reason": "A small testing_utils bug is unrelated to pytest-xdist races in captured_info.txt." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Missing Qwen3Next projections and Qwen3.5 MoE from_pretrained errors are different model-loading defects." + "reason": "Both concern Qwen3.5/MoE, but one is missing projections in layer 0 and the other is from_pretrained loading error; not the same concrete bug." }, { "left": "issue:45071", "right": "issue:47059", "accept": false, - "reason": "PretrainedConfig type checking and a CLI completion attribute error are unrelated." - }, - { - "left": "issue:44483", - "right": "issue:45464", - "accept": true, - "reason": "These appear to be the same Qwen3.5 chat/completions streaming failure, with issue:45464 describing it more specifically." + "reason": "A PretrainedConfig type-checking regression is unrelated to a cli test failure from huggingface_hub 1.22.0 attribute removal." }, { - "left": "issue:45341", + "left": "issue:45440", "right": "issue:46328", "accept": false, - "reason": "A testing_utils bug and a SequenceFeatureExtractor.pad performance issue are unrelated." + "reason": "DeepSeekV3MoE divergence and SequenceFeatureExtractor.pad inefficiency are entirely different issues." }, { "left": "issue:39401", - "right": "issue:43421", + "right": "issue:46710", "accept": false, - "reason": "Wrong offset mapping and runtime post-processor updates are different tokenizer bugs." + "reason": "Both are tokenization/model-output regressions, but the reported effects and affected models differ; no evidence of a single underlying bug." }, { - "left": "issue:44704", - "right": "issue:47110", + "left": "issue:38617", + "right": "issue:44220", "accept": false, - "reason": "Kwarg propagation in AutoProcessor.from_pretrained is unrelated to special_tokens being a list instead of a mapping." + "reason": "ImportError for layer_type_validation and _torch_extract_fbank_features behavior are unrelated." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "GenerationConfig validation failure and unrecognized model_type are different loading or validation problems." + "reason": "GLM generation config validation and OLMo-Hybrid model-type recognition are separate config-validation problems." }, { "left": "issue:43122", "right": "issue:46298", "accept": false, - "reason": "Tokenizer output changes across versions and noisy logging are unrelated." + "reason": "Different tokenization across versions and a noisy logging message are not the same bug." }, { - "left": "issue:38617", - "right": "issue:41628", + "left": "issue:44704", + "right": "issue:47110", "accept": false, - "reason": "Different import errors for different symbols; not the same missing-export bug." + "reason": "AutoProcessor kwargs forwarding and special_tokens list-vs-dict handling involve different APIs and error causes." }, { "left": "issue:38617", - "right": "issue:44220", + "right": "issue:41628", "accept": false, - "reason": "Missing layer_type_validation import and _torch_extract_fbank_features are unrelated." + "reason": "Both are import errors, but they refer to different missing symbols and different modules." }, { "left": "issue:44964", "right": "issue:45278", "accept": false, - "reason": "Phi-4 multimodal load failure and broad import errors after upgrade are too general to treat as the same concrete bug." + "reason": "Both are broad version-upgrade regressions, but one is a specific multimodal model load failure and the other is many import errors after upgrade; too generic to merge as one bug." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "One is removal of a config attribute without notice, the other is unexpected position_id keys across models; different regression classes." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 settings API behavior and FineGrainedFP8 accuracy are unrelated numeric/configuration concerns." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The verdicts are mostly conservative, but the writeup overstates the evidence for the one accepted pair. In the packet, issue:44483\u2019s visible title is only \u201c/v1/chat/completions would not accept request as usual,\u201d so saying both reports \u201cdescribe the same Qwen3.5-0.8B chat/completions streaming failure\u201d and calling it \u201cthe only true duplicate pair\u201d goes beyond what the packet directly supports. Please downgrade that language to an evidence-bounded phrasing like \u201cappears to be the same bug\u201d and avoid claiming uniqueness unless the packet explicitly establishes it." - }, - "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 }, @@ -17959,10 +18048,10 @@ "cluster_id": "cluster-10000-544", "nodes": [ "issue:27957", + "issue:29159", "issue:29925", "issue:30947", "issue:42371", - "issue:42831", "issue:43012", "issue:43066", "issue:43097", @@ -17977,6 +18066,7 @@ "issue:43874", "issue:43931", "issue:43957", + "issue:43986", "issue:44038", "issue:44075", "issue:44077", @@ -17985,21 +18075,20 @@ "issue:44261", "issue:44263", "issue:44351", - "issue:44360", "issue:44410", "issue:44484", "issue:44485", - "issue:44493", "issue:44829", "issue:44987", "issue:45071", "issue:45310", "issue:45341", + "issue:45372", "issue:45478", "issue:45542", "issue:45588", "issue:45656", - "issue:46172", + "issue:46050", "issue:46752", "issue:46762", "issue:46897", @@ -18009,251 +18098,249 @@ ], "soft_pairs": [ "issue:42371|issue:44261", - "issue:43097|issue:44493", - "issue:42371|issue:42831", "issue:44261|issue:46762", - "issue:43208|issue:46897", - "issue:43931|issue:44410", "issue:44077|issue:46981", + "issue:43931|issue:44410", + "issue:43208|issue:46897", + "issue:44038|issue:44230", + "issue:43066|issue:43122", "issue:44484|issue:45310", "issue:44075|issue:45656", - "issue:43066|issue:43122", "issue:44410|issue:45478", - "issue:44263|issue:44485", - "issue:43673|issue:45341", "issue:29925|issue:30947", - "issue:43931|issue:44038", + "issue:43673|issue:45341", "issue:43097|issue:43452", + "issue:44263|issue:44485", + "issue:43931|issue:44038", + "issue:43957|issue:44351", "issue:45071|issue:45542", + "issue:45341|issue:47448", "issue:43441|issue:44829", - "issue:44038|issue:44230", "issue:43012|issue:43577", - "issue:43957|issue:44351", - "issue:43874|issue:44261", - "issue:45341|issue:47448", - "issue:46752|issue:47116", "issue:43531|issue:45341", - "issue:27957|issue:43352", + "issue:43874|issue:44261", "issue:44117|issue:44987", - "issue:44263|issue:46172", + "issue:46752|issue:47116", "issue:43097|issue:45588", - "issue:44484|issue:45478", - "issue:44360|issue:46172" + "issue:27957|issue:43352", + "issue:43986|issue:45372", + "issue:29159|issue:30947", + "issue:43097|issue:46050" ], - "prepared_review_unit_hash": "82489b15f3499c6bb510e7bd55d143f130134867d8a2a5189e5ed2c227079b4c", + "prepared_review_unit_hash": "21d76cee8e50e6375fb331cb7b645a00a80b0d056af0239aa798ee988cfcc04d", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36508, - "estimated_input_tokens": 9127, - "estimated_eval_tokens": 18510 + "soft_pair_count": 29, + "serialized_chars": 36102, + "estimated_input_tokens": 9026, + "estimated_eval_tokens": 18308 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is thematically related to Transformers regressions and model-specific loading/runtime bugs, but the pairs are mostly not true duplicates. The soft edges are broad text matches across different models, APIs, and failure modes, so none should be merged.", - "confidence": 0.36, - "canonical_issue_reason": "issue:43097 is the cleanest representative of a concrete Transformers v5 breakage with clear user impact and multiple references; among these heterogeneous reports, it is the most broadly actionable anchor.", + "summary": "This is a mixed cluster with many unrelated Transformers issues. The only visible common thread is a loose tokenizer-regression theme, but none of the soft pairs look like the same concrete bug/change, so I would not merge any of them.", + "confidence": 0.77, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:43097 is the best single issue to represent the cluster because it describes a specific, reproducible API regression rather than a vague symptom or model-only edge case.", + "best_issue_reason": "issue:43122 is the broadest tokenizer-regression report and the best representative of the small tokenizer-related subset, but the cluster as a whole is heterogeneous.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "Both are Transformers bugs, but one is about TF32 API settings and the other about missing rms_norm_eps in MLA q_a_layernorm; different code paths and fixes." + "reason": "Both are about precision behavior, but TF32 API guidance and a missing rms_norm_eps in MLA are different problems and fixes." }, { - "left": "issue:43097", - "right": "issue:44493", + "left": "issue:44261", + "right": "issue:46762", "accept": false, - "reason": "API removal in v5 versus unexpected position-id key errors; different underlying failures." + "reason": "Both mention numerical differences, but they affect different models/code paths and do not look like one underlying bug." }, { - "left": "issue:42371", - "right": "issue:42831", + "left": "issue:44077", + "right": "issue:46981", "accept": false, - "reason": "TF32 behavior settings and FineGrainedFP8 accuracy are unrelated precision concerns, not the same bug." + "reason": "SGD args not being used and repeated 429 retries in AutoProcessor are unrelated subsystems and failures." }, { - "left": "issue:44261", - "right": "issue:46762", + "left": "issue:43931", + "right": "issue:44410", "accept": false, - "reason": "Both mention numerical issues, but they affect different models/components and do not share the same faulty code path." + "reason": "Both involve Qwen3-family loading, but weight-shape mismatch and missing projections are different concrete defects." }, { "left": "issue:43208", "right": "issue:46897", "accept": false, - "reason": "xLSTM training blockers and Florence2 double-shift are model-specific training bugs with no clear shared fix." + "reason": "xLSTM training blockers and a Florence2 double-shift loss bug are unrelated." }, { - "left": "issue:43931", - "right": "issue:44410", + "left": "issue:44038", + "right": "issue:44230", "accept": false, - "reason": "Both involve Qwen3-family models, but one is a weight-shape loading mismatch and the other missing layer projections; not the same issue." + "reason": "Both mention Qwen3-VL/MoE and fp8/FlashAttention, but one is a bug report and the other is fp8 inference support; not the same fix." }, { - "left": "issue:44077", - "right": "issue:46981", + "left": "issue:43066", + "right": "issue:43122", "accept": false, - "reason": "PatchTSMixer post_init validation and AutoProcessor 429 retry behavior are unrelated." + "reason": "These are both tokenizer regressions in v5, but one is about decoder type metadata and the other is a broad tokenization-change report; not clearly the same concrete bug." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "save_pretrained shard sizing and Qwen3.5 MoE from_pretrained loading failure are different problems." + "reason": "max_shard_size documentation/question and a Qwen3.5 MoE from_pretrained error are unrelated." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "SGD args ignored and double optimizer step under deepspeed are distinct training-loop bugs." - }, - { - "left": "issue:43066", - "right": "issue:43122", - "accept": false, - "reason": "Tokenizer decoder type regression and changed tokenization output are related only broadly; not the same underlying bug." + "reason": "Unused SGD args and double optimizer stepping with deepspeed are different optimizer bugs." }, { "left": "issue:44410", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3.5/Qwen3next-style loading issues, but the specific missing projections vs generic from_pretrained error are different." + "reason": "Both are Qwen3.5/Qwen3-next style model issues, but they concern different loading/model-definition failures." }, { - "left": "issue:44263", - "right": "issue:44485", + "left": "issue:29925", + "right": "issue:30947", "accept": false, - "reason": "A torch.split/indexer bug and a GLM-5 RoPE implementation issue are separate code paths." + "reason": "Both are tokenizer-related, but SigLIP use_fast enforcement and LLaMA3 special-token behavior are separate model-specific issues." }, { "left": "issue:43673", "right": "issue:45341", "accept": false, - "reason": "Generation cache regression and a testing_utils bug are unrelated." + "reason": "Generation cache/chunked_prefill and a testing_utils bug have no shared underlying code path." }, { - "left": "issue:29925", - "right": "issue:30947", + "left": "issue:43097", + "right": "issue:43452", "accept": false, - "reason": "Fast-tokenizer enforcement for SigLIP and special-token behavior for LLaMA3 are different tokenizer bugs." + "reason": "Tie-embeddings API removal and gguf_file loading breakage are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "A torch.split return-value bug and GLM-5 RoPE implementation are different defects in different components." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Both are Qwen3-VL related, but one is a shape mismatch on a specific checkpoint and the other is a broader MoE flash-related issue." + "reason": "Both reference Qwen3-VL/MoE, but one is a weight-shape mismatch and the other is a different from_pretrained/model-loading issue." }, { - "left": "issue:43097", - "right": "issue:43452", + "left": "issue:43957", + "right": "issue:44351", "accept": false, - "reason": "A v5 API removal and gguf_file load failures are different failure modes." + "reason": "Meta-device loading failures and a missing HybridCache import are unrelated." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "PretrainedConfig type checking and tf backend detection are unrelated issues." + "reason": "PretrainedConfig type checking and a TensorBoard-without-TF backend error are unrelated." }, { - "left": "issue:43441", - "right": "issue:44829", + "left": "issue:45341", + "right": "issue:47448", "accept": false, - "reason": "Both involve flash attention, but one is unsupported FlashAttention 2 on a model and the other is FlashAttention 3 causing degenerate training." + "reason": "A testing_utils bug and Gemma4 architecture recognition are unrelated." }, { - "left": "issue:44038", - "right": "issue:44230", + "left": "issue:43441", + "right": "issue:44829", "accept": false, - "reason": "Qwen3-VL-MoE inference regression and FP8 support request are not the same bug." + "reason": "Both mention FlashAttention, but one is an unsupported/failed model path and the other is a degenerate-training quality bug; not the same issue." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "A compile-time warning for bfloat16 and a Blip2 dtype-loading bug are different dtype issues." + "reason": "A PyTorch precision warning and a model dtype propagation bug are different problems." }, { - "left": "issue:43957", - "right": "issue:44351", + "left": "issue:43531", + "right": "issue:45341", "accept": false, - "reason": "Meta-device model loading breakage and HybridCache import failure are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Missing image-patch counting method and missing rms_norm_eps are separate multimodal/model-config defects." + "reason": "Missing image-patch counting and a normalization-epsilon precision issue are unrelated multimodal/model bugs." }, { - "left": "issue:45341", - "right": "issue:47448", + "left": "issue:44117", + "right": "issue:44987", "accept": false, - "reason": "A testing utility bug and an unrecognized architecture are clearly different." + "reason": "A tokenizer mapping None assumption and a specific loading failure for physical-intelligence/fast are not clearly the same bug." }, { "left": "issue:46752", "right": "issue:47116", "accept": false, - "reason": "Empty chat-template input crash and streaming pipeline regression are different runtime paths." + "reason": "Empty-list chat template crash and streaming pipeline regression are unrelated." }, { - "left": "issue:43531", - "right": "issue:45341", + "left": "issue:43097", + "right": "issue:45588", "accept": false, - "reason": "Qwen3-MoE sliding_window handling and testing_utils are unrelated." + "reason": "Tie-embeddings removal and a flash_attention.py sink-less crash are unrelated." }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "FlashAttention 2 support for XLMRoberta and lack of FlashAttention 2 support for NemotronH are different model-specific issues." + "reason": "Both are FlashAttention 2 related, but they concern different models and one is unsupported-model gating while the other is a specific XLM-Roberta path." }, { - "left": "issue:44117", - "right": "issue:44987", + "left": "issue:43986", + "right": "issue:45372", "accept": false, - "reason": "Tokenizer mapping returning None and a specific model load failure are not the same defect." + "reason": "AutoProcessor needing torchvision and a mistral_common ReasoningEffort import error are different processor-loading failures." }, { - "left": "issue:44263", - "right": "issue:46172", + "left": "issue:29159", + "right": "issue:30947", "accept": false, - "reason": "Different tensor-operation bugs in different models; too little evidence of a shared code path." + "reason": "Generic slow/fast tokenizer inconsistency and LLaMA3 special-token behavior are related only at a high level, not the same bug." }, { - "left": "issue:43097", - "right": "issue:45588", + "left": "issue:44261", + "right": "issue:44485", "accept": false, - "reason": "API removal regression and sink-less flash_attention crash are unrelated." + "reason": "Both discuss model math correctness, but one is a missing config epsilon and the other is RoPE implementation details; not the same fix." }, { - "left": "issue:44484", - "right": "issue:45478", + "left": "issue:45341", + "right": "issue:47448", "accept": false, - "reason": "save_pretrained shard size discussion and Qwen3.5 MoE loading failure are not duplicates." + "reason": "A testing utility issue and a Gemma4 architecture detection bug are unrelated." }, { - "left": "issue:44360", - "right": "issue:46172", + "left": "issue:43097", + "right": "issue:46050", "accept": false, - "reason": "DSA indexer ReLU omission and deepseek-v4 comb transpose issue are separate low-level bugs." + "reason": "Embedding-tie API removal and sinq save/load quantization breakage are unrelated." } ] }, @@ -18277,11 +18364,11 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-544", "nodes": [ - "issue:29159", "issue:30947", "issue:33453", "issue:34567", "issue:38617", + "issue:41950", "issue:42757", "issue:42907", "issue:43012", @@ -18294,13 +18381,13 @@ "issue:43874", "issue:43937", "issue:43986", - "issue:44038", "issue:44062", "issue:44117", "issue:44246", "issue:44261", "issue:44263", "issue:44315", + "issue:44360", "issue:44410", "issue:44484", "issue:44485", @@ -18308,11 +18395,11 @@ "issue:44936", "issue:45103", "issue:45341", - "issue:45372", + "issue:45362", "issue:45375", "issue:45440", + "issue:45478", "issue:45753", - "issue:46050", "issue:46123", "issue:46172", "issue:46298", @@ -18327,231 +18414,238 @@ "issue:46981" ], "soft_pairs": [ - "issue:43986|issue:45372", - "issue:43097|issue:46050", + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:44484|issue:45478", + "issue:43097|issue:43873", + "issue:43208|issue:43550", "issue:44261|issue:44485", - "issue:29159|issue:30947", + "issue:45440|issue:46172", "issue:42907|issue:43937", "issue:45753|issue:46369", - "issue:43097|issue:43873", - "issue:45440|issue:46172", - "issue:43208|issue:43550", "issue:43986|issue:45341", "issue:44246|issue:46981", - "issue:44410|issue:45375", "issue:44117|issue:46701", "issue:43012|issue:46821", + "issue:30947|issue:34567", "issue:44062|issue:44117", + "issue:44410|issue:45375", "issue:43673|issue:46123", "issue:30947|issue:33453", - "issue:30947|issue:34567", "issue:43874|issue:44263", - "issue:44038|issue:44410", "issue:46172|issue:46710", - "issue:46655|issue:46664", + "issue:44410|issue:45362", + "issue:44936|issue:46365", "issue:46298|issue:46752", + "issue:46655|issue:46664", "issue:44704|issue:45103", - "issue:44936|issue:46365", - "issue:38617|issue:42757", "issue:43986|issue:44484", + "issue:38617|issue:41950", + "issue:38617|issue:42757", "issue:43064|issue:44315" ], - "prepared_review_unit_hash": "f3608b33ac8481f9af9b8e155260499bd2a351fad4f52c4f1d7cb0dc194d3345", + "prepared_review_unit_hash": "fabba2cd34a740824c7877b7dfacba0b71622805d5072655e192937339a1fcef", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 35774, - "estimated_input_tokens": 8944, - "estimated_eval_tokens": 18144 + "soft_pair_count": 29, + "serialized_chars": 35954, + "estimated_input_tokens": 8989, + "estimated_eval_tokens": 18234 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No soft pair looks like the same underlying bug or change. The set is broadly heterogeneous, with only loose topical overlap (mostly tokenizer/import/model-specific regressions), so there is no PR candidate and only a weak issue-level representative.", - "confidence": 0.86, - "canonical_issue_reason": "issue:30947 is the clearest concrete tokenizer bug and the closest anchor for the tokenizer-related soft pairs, but it still does not explain the rest of the mixed cluster.", + "summary": "These items are mostly unrelated bug reports across tokenizers, model-specific generation, trainer state, quantization/offloading, and pipeline utilities. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43873 is the broadest and most actively discussed report in the set, so it is the best anchor if one issue must be chosen as representative; however, it is not actually a duplicate of the others.", "canonical_pr_reason": null, - "best_issue_reason": "issue:30947 is the best issue-level representative because it has a specific, reproducible symptom and is the most coherent among the related tokenizer reports.", + "best_issue_reason": "43873 has the widest scope and strongest ongoing discussion/inbound reference activity, making it the most suitable single issue to represent the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43986", - "right": "issue:45372", + "left": "issue:44263", + "right": "issue:46172", "accept": false, - "reason": "Different missing imports in different loading paths; same broad class of ImportError, not the same bug." + "reason": "Different bugs: DSA indexer return handling vs DeepSeek-V4 tensor cast/transpose behavior." }, { - "left": "issue:43097", - "right": "issue:46050", + "left": "issue:44360", + "right": "issue:46172", "accept": false, - "reason": "Both involve save/load or model-state behavior, but the concrete failures are unrelated." + "reason": "Unrelated code paths: missing ReLU in the DSA indexer vs DeepSeek-V4 tensor ops." }, { - "left": "issue:44261", - "right": "issue:44485", + "left": "issue:44484", + "right": "issue:45478", "accept": false, - "reason": "Both are GLM-5 related, but one is a layernorm epsilon precision issue and the other is RoPE implementation." + "reason": "One is about shard-size defaults in save_pretrained; the other is a Qwen3.5 MoE from_pretrained regression." }, { - "left": "issue:29159", - "right": "issue:30947", + "left": "issue:43097", + "right": "issue:43873", "accept": false, - "reason": "Both touch tokenizer behavior, but one is slow/fast inconsistency and the other is BOS/EOS flags not taking effect." + "reason": "Removal of a config field is unrelated to quantization/offloading failures." }, { - "left": "issue:42907", - "right": "issue:43937", + "left": "issue:43208", + "right": "issue:43550", "accept": false, - "reason": "Dequantized save failure and invalid GenerationConfig are different model-specific problems." + "reason": "xLSTM training bugs and Bamba torch.compile SDPA issues are different model/back-end problems." }, { - "left": "issue:45753", - "right": "issue:46369", + "left": "issue:44261", + "right": "issue:44485", "accept": false, - "reason": "Different models and different failure modes: infinite loop vs memory overflow." + "reason": "Both are GLM-related, but they describe different issues: layernorm epsilon config vs RoPE implementation." }, { - "left": "issue:43097", - "right": "issue:43873", + "left": "issue:45440", + "right": "issue:46172", "accept": false, - "reason": "One is a removed API/regression issue, the other is offloading with quantization; not the same code-path bug." + "reason": "Same model family but not the same defect; one is remote-implementation divergence, the other a specific tensor expression concern." }, { - "left": "issue:45440", - "right": "issue:46172", + "left": "issue:42907", + "right": "issue:43937", "accept": false, - "reason": "Both mention DeepSeek, but they concern different model implementations and different failures." + "reason": "Saving dequantized models and invalid GenerationConfig are separate failure modes." }, { - "left": "issue:43208", - "right": "issue:43550", + "left": "issue:45753", + "right": "issue:46369", "accept": false, - "reason": "xLSTM training bugs and Bamba torch.compile SDPA failure are unrelated." + "reason": "Qwen3_5 image-specific infinite loop vs SAM memory overflow are unrelated." }, { "left": "issue:43986", "right": "issue:45341", "accept": false, - "reason": "A processor-loading ImportError and a testing_utils bug are not the same underlying issue." + "reason": "Video AutoProcessor torchvision dependency error is unrelated to a testing_utils bug." }, { "left": "issue:44246", "right": "issue:46981", "accept": false, - "reason": "Both involve import/loading behavior, but one is slow imports and the other is repeated 429 retries." - }, - { - "left": "issue:44410", - "right": "issue:45375", - "accept": false, - "reason": "Missing layer projections and a missing config field are different model-definition bugs." + "reason": "Slow import and repeated 429 retry behavior are different performance/network issues." }, { "left": "issue:44117", "right": "issue:46701", "accept": false, - "reason": "Tokenizer mapping None handling and a pipeline failure are unrelated subsystems." + "reason": "Tokenizer mapping None and image-text-to-text pipeline failure are not the same bug." }, { "left": "issue:43012", "right": "issue:46821", "accept": false, - "reason": "A PyTorch bf16 compile warning and a window_size AttributeError are not the same bug." + "reason": "A bfloat16 warning during compile is unrelated to TimesFM window_size AttributeError." + }, + { + "left": "issue:30947", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer special-token handling and TrainerState token counting are different subsystems." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "AddedToken constructor misuse and tokenizer mapping assumptions are distinct tokenizer issues." + "reason": "AddedToken constructor conflict is a different tokenizer bug from mapping lookup returning None." }, { - "left": "issue:43673", - "right": "issue:46123", + "left": "issue:44410", + "right": "issue:45375", "accept": false, - "reason": "Cache handling in generation and final-batch pipeline bookkeeping are different failures." + "reason": "Missing Qwen3Next projections and missing config field in Qwen3_5MoeVisionConfig are distinct model-definition issues." }, { - "left": "issue:30947", - "right": "issue:33453", + "left": "issue:43673", + "right": "issue:46123", "accept": false, - "reason": "Both are tokenizer regressions, but one is special-token flags and the other is tokenizer loading behavior." + "reason": "Cache handling during chunked prefill is unrelated to MaskGenerationPipeline batch-finalization logic." }, { "left": "issue:30947", - "right": "issue:34567", + "right": "issue:33453", "accept": false, - "reason": "Tokenizer behavior and TrainerState token counting are unrelated." + "reason": "Both involve tokenizers, but one is LLaMA3 special-token effects and the other is a generic tokenizer-loading regression." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different GLM46V/GlmMoe code paths and different runtime errors." + "reason": "Image-patch counting in GLM46V and torch.split return handling in GlmMoeDsaIndexer are different bugs." }, { - "left": "issue:44038", - "right": "issue:44410", + "left": "issue:46172", + "right": "issue:46710", "accept": false, - "reason": "Qwen3-VL-Moe compatibility and qwen3next layer projections are separate model issues." + "reason": "A deepseek-v4 tensor-expression issue is not the same as an accuracy regression after upgrading Transformers." }, { - "left": "issue:46172", - "right": "issue:46710", + "left": "issue:44410", + "right": "issue:45362", "accept": false, - "reason": "One is a DeepSeek-V4 tensor operation issue; the other is an accuracy regression after upgrade." + "reason": "Missing projections in qwen3next and Qwen3.5 chat crashes are different model behaviors." }, { - "left": "issue:46655", - "right": "issue:46664", + "left": "issue:44936", + "right": "issue:46365", "accept": false, - "reason": "Both are SmolVLM video processor issues, but padding math and num_frames semantics are different bugs." + "reason": "trainer.evaluate after train and import failure resolved by uninstalling datasets are unrelated." }, { "left": "issue:46298", "right": "issue:46752", "accept": false, - "reason": "A noisy log message and an empty-list chat-template crash are unrelated." + "reason": "Noisy logging and apply_chat_template empty-list crash are not the same issue." + }, + { + "left": "issue:46655", + "right": "issue:46664", + "accept": false, + "reason": "Both are SmolVLM video-processor bugs, but one is padding calculation and the other is num_frames semantics; not the same concrete defect." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "Passing kwargs to cached_file and an auto_docstring AttributeError are different code paths." + "reason": "AutoProcessor kwargs passthrough and auto_docstring annotation handling are different code paths." }, { - "left": "issue:44936", - "right": "issue:46365", + "left": "issue:43986", + "right": "issue:44484", "accept": false, - "reason": "Trainer.evaluate-after-train and Trainer import broken by datasets installation are not the same failure." + "reason": "Missing torchvision handling and shard-size default behavior are unrelated." }, { "left": "issue:38617", - "right": "issue:42757", + "right": "issue:41950", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols and different packages." + "reason": "ImportError for layer_type_validation and video-classification pipeline selecting image processors are different regressions." }, { - "left": "issue:43986", - "right": "issue:44484", + "left": "issue:38617", + "right": "issue:42757", "accept": false, - "reason": "Processor loading without torchvision and save_pretrained shard sizing are unrelated." + "reason": "Both are import errors, but they refer to different missing symbols from different dependencies." }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "Wrong optimizer state handling in FSDP2/PEFT and Liger Kernel not applying with model_init are different training bugs." + "reason": "FSDP2/PEFT optimizer-state corruption and Liger Kernel not applying with model_init are separate training issues." } ] }, @@ -18577,9 +18671,8 @@ "nodes": [ "issue:30947", "issue:33357", + "issue:36010", "issue:36683", - "issue:38617", - "issue:41950", "issue:42175", "issue:42757", "issue:42890", @@ -18587,6 +18680,7 @@ "issue:43012", "issue:43064", "issue:43066", + "issue:43097", "issue:43208", "issue:43352", "issue:43502", @@ -18621,21 +18715,19 @@ "issue:46365", "issue:46421", "issue:47116", - "issue:47276", - "issue:47455" + "issue:47276" ], "soft_pairs": [ "issue:42890|issue:46298", - "issue:38617|issue:41950", - "issue:43502|issue:43992", "issue:44380|issue:46421", "issue:44451|issue:47116", + "issue:43502|issue:43992", "issue:44938|issue:46178", "issue:43208|issue:43526", "issue:43012|issue:43550", "issue:43064|issue:47116", - "issue:45712|issue:46298", "issue:43066|issue:43906", + "issue:45712|issue:46298", "issue:42907|issue:45753", "issue:42907|issue:44655", "issue:45584|issue:46365", @@ -18643,241 +18735,236 @@ "issue:43012|issue:44368", "issue:44734|issue:47116", "issue:43827|issue:47116", - "issue:43352|issue:43976", + "issue:30947|issue:33357", + "issue:36683|issue:43582", + "issue:44380|issue:45479", "issue:44077|issue:46365", + "issue:43352|issue:43976", "issue:43992|issue:44484", - "issue:30947|issue:33357", "issue:44380|issue:46132", - "issue:36683|issue:43582", "issue:44162|issue:46172", - "issue:44380|issue:45479", - "issue:42757|issue:43824", "issue:42175|issue:46365", - "issue:43012|issue:47276", + "issue:42757|issue:43824", "issue:43756|issue:44485", - "issue:46298|issue:47455", - "issue:43352|issue:44380" + "issue:43012|issue:47276", + "issue:36010|issue:42757", + "issue:43352|issue:44380", + "issue:43097|issue:46298" ], - "prepared_review_unit_hash": "d2e7287033ac51cf8246632e24e7d12ce9aa6b11e82761037d0c9163a531f114", + "prepared_review_unit_hash": "54964cc9a6e3552a5ecb73b487ecabefc8a1937ff5b4dac3d8999eafa744e12d", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 37125, - "estimated_input_tokens": 9282, - "estimated_eval_tokens": 18820 + "serialized_chars": 36581, + "estimated_input_tokens": 9146, + "estimated_eval_tokens": 18548 }, "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 that only loosely overlap on v5/regression/import/backend themes. There is no strong duplicate group; the closest candidates are still separate concrete bugs.", - "confidence": 0.84, - "canonical_issue_reason": "issue:47116 is the broadest representative here: it describes a concrete v5 runtime regression with wider user impact than the narrower, more specialized reports.", + "summary": "This is a loose semantic cluster rather than a duplicate set: the items span unrelated import errors, model-specific bugs, pipeline regressions, warnings, and docs/issues. None of the soft pairs look like the same concrete fix or a mergeable PR pair.", + "confidence": 0.9, + "canonical_issue_reason": "No single issue cleanly represents the cluster; the overlap is mostly superficial (shared Transformers/v5 vocabulary) across different code paths and models.", "canonical_pr_reason": null, - "best_issue_reason": "issue:47116 is the best single issue to stand in for the cluster because it is a clear, user-facing regression and the most general of the v5-related items.", + "best_issue_reason": "No global best issue stands out because the cluster is heterogeneous and does not center on one underlying bug.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42890", "right": "issue:46298", "accept": false, - "reason": "Test flakiness from missing set_seed is not the same as a noisy logging message." - }, - { - "left": "issue:38617", - "right": "issue:41950", - "accept": false, - "reason": "Removed import symbol error and video-classification processor lookup bug are different code paths." - }, - { - "left": "issue:43502", - "right": "issue:43992", - "accept": false, - "reason": "Local-files-only network requests and missing embed_tokens.weight in UMT5Encoder are unrelated." + "reason": "Flaky test-seed coverage vs a noisy logging message; unrelated failures." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "GPT2 attention scaling being ignored is a different bug from cache lazy init causing recompilation." + "reason": "GPT2 attention scaling ignored vs cache lazy initialization causing recompilation; different subsystems." }, { "left": "issue:44451", "right": "issue:47116", "accept": false, - "reason": "Specific model load failure and streaming pipeline regression are not the same underlying issue." + "reason": "Model loading failure vs streaming pipeline regression; no shared concrete bug." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "local_files_only making network requests vs missing embed_tokens.weight during encoder load; distinct issues." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Python 3.14 import/load failure and invalid tensorflow_text backend mapping are different failures." + "reason": "Python 3.14 import/load failure vs bad backend mapping validation; different import paths." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "xLSTM training blockers and BeitImageProcessorFast reduce_labels bug are unrelated." + "reason": "xLSTM training blockers vs BeitImageProcessorFast label reduction bug; unrelated model/components." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "bfloat16 compile warning and Bamba SDPA torch.compile failure are distinct issues." + "reason": "Compilation warning on bfloat16 vs a Bamba SDPA torch.compile failure; not the same bug." }, { "left": "issue:43064", "right": "issue:47116", "accept": false, - "reason": "FSDP2 optimizer-state corruption is not the same as the streaming pipeline regression." + "reason": "FSDP2 optimizer-state corruption vs pipeline streaming regression; different code paths." }, { - "left": "issue:45712", - "right": "issue:46298", + "left": "issue:43066", + "right": "issue:43906", "accept": false, - "reason": "Leaking dummy classes into dir(transformers) is unrelated to a noisy log message." + "reason": "Tokenizer decoder type regression vs an isolated reproduction of another issue; no shared fix." }, { - "left": "issue:43066", - "right": "issue:43906", + "left": "issue:45712", + "right": "issue:46298", "accept": false, - "reason": "Wrong tokenizer decoder type is not the same as an isolated reproduction of another issue." + "reason": "Dummy class leakage/check_repo failure vs logging noise; unrelated." }, { "left": "issue:42907", "right": "issue:45753", "accept": false, - "reason": "Save failure for dequantized Ministrals/Devstrals and Qwen3.5 image infinite loop are different bugs." + "reason": "Saving dequantized Ministrals/Devstrals vs Qwen3_5 infinite loop on one image; different bugs." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Dequantized-model save failure and Pipeline save_pretrained failure affect different object types and code paths." + "reason": "Dequantized-model save failure vs Pipeline save_pretrained behavior; different objects and failure modes." }, { "left": "issue:45584", "right": "issue:46365", "accept": false, - "reason": "Whisper empty transcription failure and Trainer import conflict with datasets are unrelated." + "reason": "Whisper empty-transcription generation bug vs Trainer import blocked by datasets; unrelated." }, { "left": "issue:43825", "right": "issue:44655", "accept": false, - "reason": "Pipeline v5 error-message behavior and saving Pipeline objects are different problems." + "reason": "Wrong translation-support error message vs inability to save Pipeline objects; both pipeline-related but not the same defect." }, { "left": "issue:43012", "right": "issue:44368", "accept": false, - "reason": "PyTorch precision warning on compile and tie_word_embeddings warning in Qwen3.5 are not the same issue." + "reason": "bfloat16 compile warning vs tie_word_embeddings warning during LoRA fine-tuning; different warnings and triggers." }, { "left": "issue:44734", "right": "issue:47116", "accept": false, - "reason": "KV-cache continuation indexing in transformers serve is separate from the streaming pipeline regression." + "reason": "KV-cache continuation tensor indexing crash in serve vs streaming pipeline regression; distinct runtime paths." }, { "left": "issue:43827", "right": "issue:47116", "accept": false, - "reason": "Docs still referencing pipeline() is a documentation issue, not the runtime streaming regression." + "reason": "Docs still referencing removed pipeline APIs vs a runtime streaming regression; docs-only versus code bug." }, { - "left": "issue:43352", - "right": "issue:43976", + "left": "issue:30947", + "right": "issue:33357", + "accept": false, + "reason": "Special-token flags not taking effect in LLaMA3 vs MacOS bus error in CLIP; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config missing vocab_size vs Apple Silicon TypeError in allocator warmup; no overlap." + }, + { + "left": "issue:44380", + "right": "issue:45479", "accept": false, - "reason": "Flash Attention 2 support gating for NemotronH is unrelated to Python version compatibility." + "reason": "Attention scaling ignored under SDPA/FlashAttention vs degenerate zero loss in sequence classification; different model behaviors." }, { "left": "issue:44077", "right": "issue:46365", "accept": false, - "reason": "patchtsmixer post_init validation and Trainer import conflict are different dependency/model issues." + "reason": "patchtsmixer post_init contract change vs Trainer import conflict with datasets; unrelated." }, { - "left": "issue:43992", - "right": "issue:44484", + "left": "issue:43352", + "right": "issue:43976", "accept": false, - "reason": "Missing encoder weights on load and max_shard_size default behavior are different save/load concerns." + "reason": "NemotronHForCausalLM FlashAttention-2 unsupported vs Python 3.9/3.10 compatibility break; different problems." }, { - "left": "issue:30947", - "right": "issue:33357", + "left": "issue:43992", + "right": "issue:44484", "accept": false, - "reason": "Special token insertion on LLaMA3 and MacOS CLIP bus error are unrelated." + "reason": "Missing encoder weight on load vs max_shard_size question; not duplicate bugs." }, { "left": "issue:44380", "right": "issue:46132", "accept": false, - "reason": "Ignored GPT2 attention scaling and AttentionInterface.register changing behavior are different attention bugs." - }, - { - "left": "issue:36683", - "right": "issue:43582", - "accept": false, - "reason": "Missing Gemma3Config vocab_size and Apple Silicon TypeError in caching_allocator_warmup are unrelated." + "reason": "GPT2 SDPA/FlashAttention scaling bug vs AttentionInterface.register behavior change; both attention-related but separate issues." }, { "left": "issue:44162", "right": "issue:46172", "accept": false, - "reason": "ESM2 being broken and deepseek-v4 tensor transpose bug are different model-specific failures." + "reason": "ESM2 broken broadly vs a deepseek-v4 dtype/transpose bug; different models and code paths." }, { - "left": "issue:44380", - "right": "issue:45479", + "left": "issue:42175", + "right": "issue:46365", "accept": false, - "reason": "Attention scaling config being ignored and degenerate zero loss in classification are unrelated." + "reason": "Package backend extras issue vs Trainer import blocked by datasets; distinct import/dependency problems." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols from different packages." + "reason": "Missing huggingface_hub symbol vs missing Transformers model class export; different imports." }, { - "left": "issue:42175", - "right": "issue:46365", + "left": "issue:43756", + "right": "issue:44485", "accept": false, - "reason": "Missing TensorFlow backend from torch install and Trainer import conflict with datasets are separate dependency problems." + "reason": "Smollm3 RoPE layer-count bug vs GLM-5 RoPE implementation discussion; different models and failure modes." }, { "left": "issue:43012", "right": "issue:47276", "accept": false, - "reason": "A compile precision warning is not the same as a misleading synced_gpus warning message." - }, - { - "left": "issue:43756", - "right": "issue:44485", - "accept": false, - "reason": "Smollm3 RoPE layer count mismatch and GLM-5 RoPE implementation discussion are not the same concrete bug." + "reason": "Compile warning on bfloat16 vs misleading synced_gpus warning; only superficially similar as warnings." }, { - "left": "issue:46298", - "right": "issue:47455", + "left": "issue:36010", + "right": "issue:42757", "accept": false, - "reason": "Noisy logging and kernel version checking are unrelated." + "reason": "Missing GenerationMixin import vs missing is_offline_mode import; separate broken symbols." }, { - "left": "issue:43352", - "right": "issue:44380", + "left": "issue:43097", + "right": "issue:46298", "accept": false, - "reason": "Flash Attention 2 unsupported warning and GPT2 attention scaling being ignored are different issues." + "reason": "Removed tie_embeddings_and_encoder_decoder API vs a noisy logging message; unrelated." } ] }, @@ -18904,11 +18991,10 @@ "issue:29870", "issue:29925", "issue:33453", - "issue:36010", "issue:36296", "issue:36683", "issue:38617", - "issue:42757", + "issue:42548", "issue:42831", "issue:42890", "issue:43064", @@ -18916,6 +19002,7 @@ "issue:43352", "issue:43475", "issue:43502", + "issue:43525", "issue:43611", "issue:43618", "issue:43761", @@ -18925,6 +19012,7 @@ "issue:44222", "issue:44261", "issue:44263", + "issue:44351", "issue:44355", "issue:44380", "issue:44483", @@ -18932,7 +19020,6 @@ "issue:44625", "issue:44655", "issue:44855", - "issue:44908", "issue:44938", "issue:45216", "issue:45276", @@ -18943,274 +19030,266 @@ "issue:46202", "issue:46365", "issue:46512", - "issue:46851", "issue:46877", "issue:47059", - "issue:47110", + "issue:47276", "issue:47321", "issue:47332" ], "soft_pairs": [ - "issue:36010|issue:42757", "issue:42831|issue:44261", - "issue:44222|issue:45478", + "issue:43618|issue:43761", "issue:44483|issue:44514", "issue:43064|issue:44380", + "issue:44222|issue:45478", "issue:42890|issue:47059", - "issue:43618|issue:43761", "issue:44625|issue:46365", "issue:45276|issue:46077", - "issue:38617|issue:43502", - "issue:44655|issue:47332", "issue:44655|issue:47321", + "issue:44655|issue:47332", + "issue:38617|issue:43502", "issue:44625|issue:46077", - "issue:43299|issue:43867", "issue:43611|issue:44655", - "issue:43618|issue:44380", + "issue:43299|issue:43867", "issue:36683|issue:43475", "issue:43352|issue:45593", + "issue:44380|issue:47276", "issue:36683|issue:44077", "issue:46077|issue:46512", "issue:44263|issue:45759", - "issue:43976|issue:44938", "issue:44355|issue:46877", + "issue:43976|issue:44938", "issue:29925|issue:33453", "issue:43867|issue:45216", "issue:36683|issue:44625", "issue:44938|issue:47059", "issue:43618|issue:45216", - "issue:36683|issue:47110", + "issue:36683|issue:43525", "issue:44355|issue:44855", + "issue:42548|issue:44351", "issue:29870|issue:36296", - "issue:43867|issue:46202", - "issue:44908|issue:46851" + "issue:43867|issue:46202" ], - "prepared_review_unit_hash": "f4e53d7ae5a2446acb75544472fa40935d580934d2b00093761a0fe3420a56b0", + "prepared_review_unit_hash": "e6f57b618664f93565ba75a8faa610c36e13a5bb14bb30679be4c1a2f04a879a", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37395, - "estimated_input_tokens": 9349, - "estimated_eval_tokens": 18954 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36504, + "estimated_input_tokens": 9126, + "estimated_eval_tokens": 18508 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous. The soft pairs mostly share broad vocabulary like import/load/save/regression, but they describe different concrete bugs in different code paths, so I rejected all candidate duplicates. No single issue cleanly represents the whole set.", - "confidence": 0.9, - "canonical_issue_reason": "No canonical issue: the items span unrelated failures across imports, model loading/saving, tokenizers, configs, tests, and training; the overlap is only topical, not duplicate-level.", + "summary": "This cluster is heterogeneous: it mixes unrelated model/config regressions, import errors, tokenizer issues, training bugs, and API behavior changes. There is no single duplicate theme strong enough to treat any item as a canonical cluster anchor, and none of the soft pairs look like the same underlying bug/change. No PRs are present in the set.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items span multiple unrelated subsystems and failure modes, so selecting one issue would be arbitrary rather than duplicate-driven.", "canonical_pr_reason": null, - "best_issue_reason": "No strong global issue anchor. If forced, a narrowly scoped regression report would be preferable, but none of these subsumes the rest.", + "best_issue_reason": "If a representative issue is needed, issue 36683 is the strongest standalone exemplar because it is a concrete, reproducible runtime bug report with substantial discussion activity; however, it is not representative enough to serve as a canonical duplicate anchor for this mixed cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:36010", - "right": "issue:42757", - "accept": false, - "reason": "Both are ImportErrors, but for different symbols and modules (`GenerationMixin` vs `is_offline_mode`)." - }, { "left": "issue:42831", "right": "issue:44261", "accept": false, - "reason": "Both mention accuracy/precision, but one is a generic FP8 accuracy issue and the other is a specific MLA RMS norm epsilon mismatch." + "reason": "Different problems: FineGrainedFP8 accuracy vs MLA q_a_layernorm missing rms_norm_eps. Both are precision-related, but the underlying bugs and fixes are not the same." }, { - "left": "issue:44222", - "right": "issue:45478", + "left": "issue:43618", + "right": "issue:43761", "accept": false, - "reason": "Save-pretrained vs from-pretrained loading are different code paths and failure modes." + "reason": "Related CLIP area, but one is missing attentions assignment and the other is hidden_states being None in forward. Different outputs, different code paths." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "Both involve chat/input handling, but one is an API request rejection and the other is a processor crash on batched padding=False input." + "reason": "Different request-handling bugs: chat completions API rejection vs batched Qwen2.5-VL chat template crash with padding=False." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "Different subsystems: distributed trainer optimizer state handling vs GPT-2 attention scaling under SDPA/FlashAttention." + "reason": "Different training/runtime issues: optimizer state corruption under FSDP2+PEFT vs GPT2 attention scaling ignored under SDPA/FlashAttention." }, { - "left": "issue:42890", - "right": "issue:47059", + "left": "issue:44222", + "right": "issue:45478", "accept": false, - "reason": "Both are test failures, but caused by different missing setup / CLI compatibility issues." + "reason": "Both mention FP8/MoE, but one is save_pretrained for FP8 moe models and the other is from_pretrained failure for Qwen3.5 MoE. Not the same concrete bug." }, { - "left": "issue:43618", - "right": "issue:43761", + "left": "issue:42890", + "right": "issue:47059", "accept": false, - "reason": "Both are CLIP regressions, but one is missing attentions output and the other is hidden_states not returned." + "reason": "One is flaky SAM HQ tests due to missing set_seed; the other is a CLI incompatibility with huggingface_hub 1.22.0. Different failure modes." }, { "left": "issue:44625", "right": "issue:46365", "accept": false, - "reason": "Qwen3.5 config propagation bug vs an unrelated Trainer import conflict with `datasets`." + "reason": "Qwen3.5 num_labels propagation bug vs inability to import Trainer until datasets is uninstalled. Unrelated subsystems." }, { "left": "issue:45276", "right": "issue:46077", "accept": false, - "reason": "Different Gemma4 bugs: embedding resizing vs bidirectional attention mask construction." + "reason": "Both are Gemma4-related, but one is resize_token_embeddings not updating embeddings and the other is bidirectional attention still building causal masks. Different code paths." }, { - "left": "issue:38617", - "right": "issue:43502", + "left": "issue:44655", + "right": "issue:47321", "accept": false, - "reason": "An import error in `configuration_utils` is unrelated to unwanted network requests with `local_files_only=True`." + "reason": "Pipeline save_pretrained failure vs AMD quark class not updated. No shared underlying bug." }, { "left": "issue:44655", "right": "issue:47332", "accept": false, - "reason": "Both concern saving, but one is pipeline serialization and the other is disk-offloaded Qwen3-VL checkpoint saving." + "reason": "Pipeline save_pretrained failure vs inability to save disk-offloaded Qwen3-VL. Different save/offload behavior and likely different fixes." }, { - "left": "issue:44655", - "right": "issue:47321", + "left": "issue:38617", + "right": "issue:43502", "accept": false, - "reason": "Pipeline save failure is unrelated to an AMD Quark class update issue." + "reason": "Both involve import/loading behavior, but one is a missing symbol import and the other is local_files_only still making API requests. Not the same issue." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "Different Qwen3.5 vs Gemma4 configuration bugs with no shared failing path." - }, - { - "left": "issue:43299", - "right": "issue:43867", - "accept": false, - "reason": "Both are load-model regressions, but the reported symptoms and likely root causes differ (`base_model_prefix` vs sorted state_dict)." + "reason": "Qwen3.5 num_labels propagation and Gemma4 bidirectional attention are distinct configuration bugs in different model families." }, { "left": "issue:43611", "right": "issue:44655", "accept": false, - "reason": "Model loading with `base_model_prefix` is unrelated to saving Pipeline objects." + "reason": "Base_model_prefix loading regression is unrelated to Pipeline save_pretrained behavior." }, { - "left": "issue:43618", - "right": "issue:44380", + "left": "issue:43299", + "right": "issue:43867", "accept": false, - "reason": "Attention-related wording only; one is missing attentions output, the other ignores attention scaling config." + "reason": "Qwen3VL MoE loading regression from v5 vs sorted state_dict load error. Both are loading-related, but not the same code-path problem." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Gemma3 vocab_size config bug and SAM 3 Video missing `fpn_position_embeddings` are unrelated model attribute errors." + "reason": "Gemma3Config missing vocab_size is unrelated to SAM 3 Video missing fpn_position_embeddings." }, { "left": "issue:43352", "right": "issue:45593", "accept": false, - "reason": "FlashAttention support gating and D-FINE auxiliary-loss logic are different bugs." + "reason": "Flash Attention 2 support check for NemotronHForCausalLM vs D-FINE auxiliary losses disabled when denoising is off. Completely different bugs." + }, + { + "left": "issue:44380", + "right": "issue:47276", + "accept": false, + "reason": "GPT2 attention scaling ignored in backends vs misleading synced_gpus warning. One is functional behavior, the other is warning text." }, { "left": "issue:36683", "right": "issue:44077", "accept": false, - "reason": "A missing config attribute is unrelated to the `patchtsmixer` `post_init` policy change." + "reason": "Gemma3Config missing vocab_size and patchtsmixer post_init validation are unrelated model/config issues." }, { "left": "issue:46077", "right": "issue:46512", "accept": false, - "reason": "Bidirectional attention mask construction and stale `config.dtype` are separate config bugs." + "reason": "Gemma4 attention mask bug vs from_config(dtype=...) leaving model.config.dtype stale. Different configuration/state propagation bugs." }, { "left": "issue:44263", "right": "issue:45759", "accept": false, - "reason": "Different failures: `torch.split` behavior in a MoE indexer vs unwrapping `text_config` for composite multimodal configs." + "reason": "torch.split return values in GlmMoeDsaIndexer vs AutoModelForCausalLM.from_config not unwrapping text_config for composite Qwen configs. Different model-loading logic." }, { - "left": "issue:43976", - "right": "issue:44938", + "left": "issue:44355", + "right": "issue:46877", "accept": false, - "reason": "Both are Python-version compatibility reports, but for different versions and failure modes." + "reason": "Compiled Python file import errors vs auto_docstring printing doc-lint diagnostics at import time. Not the same underlying issue." }, { - "left": "issue:44355", - "right": "issue:46877", + "left": "issue:43976", + "right": "issue:44938", "accept": false, - "reason": "Compiled Python import errors and stdout docstring lint noise are unrelated." + "reason": "Python 3.9+ compatibility problem in Transformers 5.1.0 vs Python 3.14 import/load failure. Similar platform topic, but not the same bug." }, { "left": "issue:29925", "right": "issue:33453", "accept": false, - "reason": "Tokenizer fast-mode enforcement is not the same as a broad tokenizer-loading regression." + "reason": "SigLIP use_fast enforcement vs tokenizer loading regression. The latter is broader and not clearly the same as the former." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Load-time state_dict ordering problems and save_pretrained checkpoint corruption are different sides of the API and different bugs." + "reason": "State_dict sorting load error vs Qwen3.5 saved checkpoint regression. Both touch save/load, but the failing mechanism differs." }, { "left": "issue:36683", "right": "issue:44625", "accept": false, - "reason": "Different config fields (`vocab_size` vs `num_labels`) and different affected models." + "reason": "Gemma3Config vocab_size missing vs Qwen3.5 num_labels propagation. Different configs and different missing fields." }, { "left": "issue:44938", "right": "issue:47059", "accept": false, - "reason": "Python 3.14 import failure vs CLI breakage with huggingface_hub 1.22.0 are unrelated compatibility issues." + "reason": "Python 3.14 load failure vs CLI breakage with huggingface_hub 1.22.0. Unrelated compatibility issues." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "Missing CLIP attentions output is unrelated to Qwen3.5 save_pretrained checkpoint correctness." + "reason": "CLIP attentions regression vs Qwen3.5 checkpoint save regression. No shared code-path or fix." }, { "left": "issue:36683", - "right": "issue:47110", + "right": "issue:43525", "accept": false, - "reason": "`Gemma3Config.vocab_size` missing and `special_tokens` being a list instead of a dict are unrelated." + "reason": "Gemma3Config missing vocab_size vs Llama4Config missing pad_token_id. Both are missing-config-field errors, but for different models and different fields." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Both are import-related, but the causes differ: compiled file handling vs a Python 3.13 `torch.jit.script` parsing issue." + "reason": "Compiled Python import issues vs torch.jit.script parsing failure on Python 3.13. Different causes and fixes." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "PreTrainedModel import failure vs HybridCache import failure. Same broad import category, but different missing symbols and likely different breakages." }, { "left": "issue:29870", "right": "issue:36296", "accept": false, - "reason": "BART generate decoder mask initialization and tensor-parallel training optimizer state corruption are unrelated." + "reason": "BART generate() decoder_attention_mask bug vs tensor parallel training bug. Completely different model/runtime paths." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "A model loading error when sorting state_dict is not the same as Gemma pipeline roundtrip tokenization failure." - }, - { - "left": "issue:44908", - "right": "issue:46851", - "accept": false, - "reason": "Scheduler kwargs handling and Qwen3.5 variable-length position_ids are separate bugs in different subsystems." + "reason": "state_dict sorting load error vs Gemma pipeline roundtrip failure on \u2581. Different loading/serialization issues." } ] }, @@ -19256,12 +19335,12 @@ "issue:44077", "issue:44112", "issue:44222", + "issue:44230", "issue:44246", - "issue:44351", "issue:44360", "issue:44393", "issue:44492", - "issue:44734", + "issue:44908", "issue:44964", "issue:45125", "issue:45278", @@ -19276,7 +19355,7 @@ "issue:46285", "issue:46291", "issue:46365", - "issue:46617", + "issue:46851", "issue:47000", "issue:47317", "issue:47337", @@ -19284,24 +19363,24 @@ ], "soft_pairs": [ "issue:44393|issue:47337", + "issue:44908|issue:46851", "issue:43992|issue:47317", "issue:43976|issue:47455", - "issue:44075|issue:46365", "issue:44393|issue:47000", + "issue:44075|issue:46365", "issue:43010|issue:45278", - "issue:43992|issue:45310", "issue:29870|issue:33453", - "issue:42548|issue:44351", + "issue:43992|issue:45310", "issue:36296|issue:43575", - "issue:43064|issue:44492", "issue:46291|issue:47455", "issue:43618|issue:44077", - "issue:44222|issue:45125", + "issue:43064|issue:44492", "issue:41669|issue:44246", "issue:42890|issue:45561", - "issue:43867|issue:45375", - "issue:43867|issue:45357", "issue:28282|issue:33453", + "issue:43867|issue:45357", + "issue:44222|issue:45125", + "issue:43867|issue:45375", "issue:43010|issue:43976", "issue:43618|issue:46285", "issue:46234|issue:47455", @@ -19311,237 +19390,237 @@ "issue:44360|issue:45440", "issue:43502|issue:44062", "issue:42890|issue:44112", + "issue:41628|issue:42548", "issue:43525|issue:47317", - "issue:42890|issue:44964", - "issue:41628|issue:47455", - "issue:44734|issue:46617" + "issue:44222|issue:44230", + "issue:42890|issue:44964" ], - "prepared_review_unit_hash": "2be1d05b4b414119a1f762b618f374481cfde84cd52db738edf9197cd792fef0", + "prepared_review_unit_hash": "d47cdb39ed00475c2e3a0b37329b06366a8cf4ed3d385b09ec3d9f61e2a0ab1d", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 36301, - "estimated_input_tokens": 9076, - "estimated_eval_tokens": 18408 + "serialized_chars": 36265, + "estimated_input_tokens": 9067, + "estimated_eval_tokens": 18390 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is fairly heterogeneous: most items are distinct import/model/config/training bugs, not true duplicates. I\u2019d only merge two soft pairs as same underlying issues, and there are no PRs here.", - "confidence": 0.81, - "canonical_issue_reason": "issue:45278 is the broadest representative of the import-breakage subgroup, with a high-level symptom that overlaps several of the import-related reports in this set.", + "summary": "Most pairs are related only by broad subsystem similarity, not the same concrete bug. I would deduplicate two pairs only: the Qwen3.5 MoE from_pretrained regression and the import-slowdown reports. No PRs are present.", + "confidence": 0.64, + "canonical_issue_reason": "issue:41669 is the strongest anchor in the set: it describes a concrete import-time regression with an explicit suspected cause, and it cleanly absorbs the matching slowdown report in issue:44246.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 has the clearest cluster-wide scope and the best chance of representing the largest shared symptom family among these issues.", + "best_issue_reason": "issue:41669 is the most actionable and specific issue overall; it has a clear symptom, a likely root cause, and a plausible duplicate counterpart.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44393", "right": "issue:47337", "accept": false, - "reason": "Both mention bbox values, but one is model output hallucination and the other is a flaky test about coordinate range; not the same bug." + "reason": "Both mention bbox/output problems, but one is a model output issue and the other is a flaky test; they do not look like the same underlying bug." + }, + { + "left": "issue:44908", + "right": "issue:46851", + "accept": false, + "reason": "Scheduler kwargs handling and Qwen3.5 variable-length position_ids are unrelated bugs." }, { "left": "issue:43992", "right": "issue:47317", "accept": false, - "reason": "UMT5 encoder weight loading and PPFormulaNet training loss are unrelated code paths." + "reason": "UMT5Encoder missing weights and a PPFormulaNet training-loss bug are different code paths." }, { "left": "issue:43976", "right": "issue:47455", "accept": false, - "reason": "Python-version compatibility and kernel-version checking are different failure modes." - }, - { - "left": "issue:44075", - "right": "issue:46365", - "accept": false, - "reason": "Unused SGD args and a Trainer import failure are unrelated issues." + "reason": "A Python version compatibility issue and a kernel version-check bug are not the same defect." }, { "left": "issue:44393", "right": "issue:47000", "accept": false, - "reason": "BBox hallucination in Qwen3-VL and RTDetr loss infeasibility are different model bugs." + "reason": "Both are vision/detection-related, but one is about Qwen3-VL bbox output and the other is RTDetrLoss infeasibility; different failures." }, { - "left": "issue:43010", - "right": "issue:45278", + "left": "issue:44075", + "right": "issue:46365", "accept": false, - "reason": "no_grad decoration for cache updates is unrelated to import errors after a version update." + "reason": "Optimizer SGD args being ignored and Trainer import failing with datasets are unrelated." }, { - "left": "issue:43992", - "right": "issue:45310", + "left": "issue:43010", + "right": "issue:45278", "accept": false, - "reason": "Missing encoder weights and Qwen3.5 MoE from_pretrained failure are different problems." + "reason": "Cache update no_grad decoration and broad import errors after upgrade are different problems." }, { "left": "issue:29870", "right": "issue:33453", "accept": false, - "reason": "Decoder attention-mask use-before-assignment and tokenizer loading regression are not the same bug." + "reason": "BART generate() decoder_attention_mask and tokenizer loading regression are separate issues." }, { - "left": "issue:42548", - "right": "issue:44351", + "left": "issue:43992", + "right": "issue:45310", "accept": false, - "reason": "Both are import errors, but they concern different missing symbols and likely different causes." + "reason": "Missing embed_tokens.weight in UMT5Encoder is not the same as Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:36296", "right": "issue:43575", "accept": false, - "reason": "A tensor-parallel training bug and a tp-related OOM on a specific model are not the same fix." - }, - { - "left": "issue:43064", - "right": "issue:44492", - "accept": false, - "reason": "Optimizer state corruption and a cache-strategy typo are unrelated." + "reason": "Tensor-parallel training state corruption and an OOM while loading a large model are different failure modes." }, { "left": "issue:46291", "right": "issue:47455", "accept": false, - "reason": "Both involve kernels/versioning, but the concrete failure and likely fix are not clearly the same." + "reason": "Importing pipelines with kernels v0.15 and a generic kernel version-check bug are related only at a very high level." }, { "left": "issue:43618", "right": "issue:44077", "accept": false, - "reason": "CLIPOutput attention assignment and patchtsmixer post_init behavior are unrelated." + "reason": "Missing CLIPOutput attentions and patchtsmixer post_init gating are not the same bug." }, { - "left": "issue:44222", - "right": "issue:45125", + "left": "issue:43064", + "right": "issue:44492", "accept": false, - "reason": "FP8 save_pretrained behavior and missing _tp_plan for Qwen3_5Moe are different code paths." + "reason": "FSDP2/PEFT optimizer state corruption and a cache-strategy typo are unrelated." }, { "left": "issue:41669", "right": "issue:44246", "accept": true, - "reason": "Both describe the same import-time slowdown, and the import-* removal matches the reported intermittent long import symptom." + "reason": "Both describe the same import-time slowdown in transformers; 41669 gives the concrete cause (import * usage), while 44246 is the symptom report." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "A missing set_seed flaky test and an xdist race on captured_info.txt are different CI issues." + "reason": "A flaky SAM HQ test due to missing set_seed() is unrelated to an xdist race on captured_info.txt." }, { - "left": "issue:43867", - "right": "issue:45375", + "left": "issue:28282", + "right": "issue:33453", "accept": false, - "reason": "State-dict loading order and a missing vision-config field are unrelated." + "reason": "Missing PyTorch for AutoModel and tokenizer loading regression are not the same underlying issue." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "State-dict loading failure and incorrect visual encoder keys in save_pretrained are different issues." + "reason": "State-dict loading sorted error and incorrect visual encoder keys on save_pretrained are different model-save/load bugs." }, { - "left": "issue:28282", - "right": "issue:33453", + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained behavior and a missing tensor-parallel plan for Qwen3_5Moe are separate defects." + }, + { + "left": "issue:43867", + "right": "issue:45375", "accept": false, - "reason": "PyTorch-missing ImportError and tokenizer-loading regression are unrelated." + "reason": "Sorted state_dict loading and missing deepstack_visual_indexes are unrelated." }, { "left": "issue:43010", "right": "issue:43976", "accept": false, - "reason": "Cache no_grad decoration and Python-version support are unrelated." + "reason": "Cache update decoration and Python version compatibility are unrelated." }, { "left": "issue:43618", "right": "issue:46285", "accept": false, - "reason": "Both touch CLIP, but they are different structural/API regressions." + "reason": "CLIPOutput attentions and CLIPTextModel structure inconsistency are different CLIP-family issues." }, { "left": "issue:46234", "right": "issue:47455", "accept": false, - "reason": "Missing import_utils helper and kernel version checking are not the same bug." + "reason": "Missing import_utils symbol and kernel version-check logic are not the same bug." }, { "left": "issue:45125", "right": "issue:45478", "accept": true, - "reason": "These report the same Qwen3.5-MoE from_pretrained/tp-plan failure, with nearly identical symptom and scope." + "reason": "These are the same Qwen3.5 MoE from_pretrained error, with near-identical titles and the same model path." }, { "left": "issue:43010", "right": "issue:45850", "accept": false, - "reason": "Cache-update decoration and broad version-breakage complaints are too different and unspecific." + "reason": "A cache/Layer no_grad concern and a broad complaint that 5.0 breaks too many models are not the same concrete bug." }, { "left": "issue:45125", "right": "issue:45375", "accept": false, - "reason": "Missing tp plan and missing deepstack_visual_indexes are different Qwen3.5 config issues." + "reason": "Both involve Qwen3.5 MoE vision components, but one is a from_pretrained error and the other is a missing config field." }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "A ReLU discussion in the DSA indexer and DeepSeekV3 implementation divergence are unrelated." + "reason": "A DSA indexer activation bug and divergence of DeepseekV3MoE from remote implementation are unrelated." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "local_files_only network leakage and AddedToken argument collision are unrelated." + "reason": "Offline-loading API requests and AddedToken special-argument duplication are different code paths." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "SamHQ seed flakiness and GraniteSpeech stale device override failures are different CI bugs." + "reason": "A flaky integration test and a CI device-override race are not the same failure." }, { - "left": "issue:43525", - "right": "issue:47317", + "left": "issue:41628", + "right": "issue:42548", "accept": false, - "reason": "Llama4Config missing pad_token_id and PPFormulaNet training loss are unrelated." + "reason": "Missing AutoImageProcessor and missing PreTrainedModel are both import/export regressions, but they are different symbols and likely different fixes." }, { - "left": "issue:42890", - "right": "issue:44964", + "left": "issue:43525", + "right": "issue:47317", "accept": false, - "reason": "SamHQ flaky tests and Phi-4-multimodal loading failure are unrelated." + "reason": "Llama4Config missing pad_token_id and a PPFormulaNet training-loss bug are unrelated." }, { - "left": "issue:41628", - "right": "issue:47455", + "left": "issue:44222", + "right": "issue:44230", "accept": false, - "reason": "AutoImageProcessor import failure and kernel version checking are different import issues." + "reason": "FP8 save_pretrained for MoE and fp8 inference support for Qwen3-VL/Qwen3.5 MoE are different requests." }, { - "left": "issue:44734", - "right": "issue:46617", + "left": "issue:42890", + "right": "issue:44964", "accept": false, - "reason": "KV-cache continuation crash in serve and cache FileNotFoundError on trust_remote_code loading are unrelated." + "reason": "SAM HQ test flakiness and Phi-4 multimodal model loading are unrelated issues." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet, and the soft-edge calls are conservative. The two accepted pairs look plausibly duplicate or same-bug issues, while the rejections avoid overstating weak similarity. No PR-pair mergeability issues apply here." + "feedback": "The verdicts are grounded in the packet and stay conservative. The two accepted duplicates are well supported by the titles: the Qwen3.5 MoE from_pretrained reports (45125/45478) are effectively the same issue, and the import-slowdown reports (41669/44246) plausibly describe the same regression/symptom pair. The rejected pairs are mostly limited to broad subsystem similarity, which is appropriately cautious." }, "evaluator_used": true, "retried": false, @@ -19567,9 +19646,10 @@ "issue:30990", "issue:31795", "issue:36296", + "issue:36683", + "issue:38175", "issue:38617", "issue:41628", - "issue:42548", "issue:42757", "issue:42898", "issue:43010", @@ -19578,21 +19658,20 @@ "issue:43299", "issue:43475", "issue:43502", + "issue:43526", "issue:43576", "issue:43723", "issue:43824", "issue:43827", "issue:43874", - "issue:43901", "issue:44038", "issue:44077", - "issue:44222", "issue:44230", - "issue:44351", "issue:44355", "issue:44387", "issue:44485", "issue:44556", + "issue:44734", "issue:44749", "issue:44861", "issue:44961", @@ -19608,283 +19687,278 @@ "issue:46726", "issue:47059", "issue:47110", - "issue:47317" + "issue:47317", + "issue:47455" ], "soft_pairs": [ "issue:44387|issue:44556", - "issue:41628|issue:42548", + "issue:44734|issue:46617", + "issue:41628|issue:47455", + "issue:46143|issue:46144", "issue:43723|issue:44556", - "issue:44222|issue:44230", "issue:43010|issue:43065", "issue:43824|issue:46617", "issue:43576|issue:46132", - "issue:45071|issue:47317", "issue:42898|issue:46664", - "issue:46143|issue:46144", - "issue:43502|issue:44485", + "issue:45071|issue:47317", "issue:44355|issue:44749", - "issue:43874|issue:46617", + "issue:43502|issue:44485", "issue:30990|issue:31795", - "issue:45850|issue:47110", + "issue:43874|issue:46617", "issue:43010|issue:43232", - "issue:43502|issue:44222", + "issue:45850|issue:47110", + "issue:44230|issue:46095", "issue:43299|issue:44556", + "issue:38175|issue:43502", + "issue:43827|issue:44961", "issue:46664|issue:46726", "issue:30333|issue:36296", - "issue:43827|issue:44961", "issue:29870|issue:31795", "issue:44038|issue:45125", - "issue:44230|issue:46095", "issue:43576|issue:43824", - "issue:43502|issue:44961", "issue:44355|issue:45071", "issue:43576|issue:45071", + "issue:43502|issue:44961", "issue:43576|issue:47059", - "issue:42757|issue:43576", - "issue:36296|issue:43901", + "issue:36296|issue:43526", "issue:31795|issue:38617", + "issue:42757|issue:43576", + "issue:36683|issue:47110", "issue:43475|issue:46617", - "issue:44077|issue:44861", - "issue:44351|issue:46617" + "issue:44077|issue:44861" ], - "prepared_review_unit_hash": "7dcd31e869a6c378c2039990146d2e0fc5283bc5ccef037f4f239a99aaf018d8", + "prepared_review_unit_hash": "2172d0da17f90db1cf9f302d4734709d7041abc9ea5e48fe9c3cf70b80c1acaf", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 35, - "serialized_chars": 37206, - "estimated_input_tokens": 9302, - "estimated_eval_tokens": 18860 + "serialized_chars": 37870, + "estimated_input_tokens": 9468, + "estimated_eval_tokens": 19192 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are mostly a loose v5-regression cluster with mixed model-loading, import, CLI, docs, and model-specific bugs; none of the soft pairs look like true duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "issue:44556 is the clearest representative regression report: a concrete v5 upgrade/reload failure with broad applicability.", + "summary": "This set is a loose collection of Transformers-related regressions, import errors, and docs issues, but the soft pairs are mostly different bugs and should not be merged as duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:45850 is the broadest umbrella report for the general \u201cv5 breaks many models\u201d regression theme.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44556 is the best umbrella candidate in this set because it is concrete and broadly framed, unlike the many narrowly scoped import/docs/model-specific reports.", + "best_issue_reason": "issue:44556 is the most concrete and reproducible of the v5-loading regression reports, so it is the best representative anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Different failure modes: int4 quantization OOM vs checkpoint reload breakage after v5 upgrade." + "reason": "Different failure modes: int4 quantization memory/OOM vs checkpoint reload compatibility after upgrade." + }, + { + "left": "issue:44734", + "right": "issue:46617", + "accept": false, + "reason": "Serve KV-cache indexing crash is unrelated to trust_remote_code cache loading FileNotFoundError." }, { "left": "issue:41628", - "right": "issue:42548", + "right": "issue:47455", "accept": false, - "reason": "Both are import errors, but for different symbols and likely different breakpoints (`AutoImageProcessor` vs `PreTrainedModel`)." + "reason": "Importing AutoImageProcessor is unrelated to a kernel version check bug." }, { - "left": "issue:43723", - "right": "issue:44556", + "left": "issue:46143", + "right": "issue:46144", "accept": false, - "reason": "Tokenizer loading in v5 is not the same bug as failing to reload checkpoints after an upgrade." + "reason": "Both concern RoFormer, but one is kwargs passthrough and the other is attention-interface integration; not the same bug." }, { - "left": "issue:44222", - "right": "issue:44230", + "left": "issue:43723", + "right": "issue:44556", "accept": false, - "reason": "FP8 save_pretrained bug vs FP8 inference support request; related area, not the same bug." + "reason": "Tokenizer loading in v5 is a different code path from reloading old checkpoints after upgrade." }, { "left": "issue:43010", "right": "issue:43065", "accept": false, - "reason": "Cache/layer `update()` no_grad issue is unrelated to the dummy Conv2d in Sam3PixelDecoder." + "reason": "Cache no_grad decoration and Sam3PixelDecoder dummy Conv2d are unrelated changes." }, { "left": "issue:43824", "right": "issue:46617", "accept": false, - "reason": "Specific Qwen2.5-VL import error vs generic trust_remote_code cache FileNotFoundError." + "reason": "Missing Qwen2_5_VL export and loading a trust_remote_code model from cache are different import/load problems." }, { "left": "issue:43576", "right": "issue:46132", "accept": false, - "reason": "Broken `transformers env` command is not the same as `AttentionInterface.register` changing registered-function behavior." + "reason": "CLI env command breakage is unrelated to AttentionInterface.register behavior changes." }, { - "left": "issue:45071", - "right": "issue:47317", + "left": "issue:42898", + "right": "issue:46664", "accept": false, - "reason": "PretrainedConfig type-checking regression vs PPFormulaNet training-loss bug; no shared codepath." + "reason": "Tokenization-space cleanup behavior and video num_frames semantics are unrelated." }, { - "left": "issue:42898", - "right": "issue:46664", + "left": "issue:45071", + "right": "issue:47317", "accept": false, - "reason": "Tokenization-space behavior change and video `num_frames` semantics are unrelated." + "reason": "PretrainedConfig type checking and PPFormulaNet training-loss bug are different issues." }, { - "left": "issue:46143", - "right": "issue:46144", + "left": "issue:44355", + "right": "issue:44749", "accept": false, - "reason": "Both concern RoFormer, but one is missing `**kwargs` passthrough and the other says attention doesn't use the interface; not one concrete bug." + "reason": "Compiled Python file errors are unrelated to preprocessing slowdown after version upgrade." }, { "left": "issue:43502", "right": "issue:44485", "accept": false, - "reason": "`local_files_only=True` still triggering requests is unrelated to GLM-5 RoPE implementation." + "reason": "local_files_only network requests and GLM-5 RoPE implementation are unrelated." }, { - "left": "issue:44355", - "right": "issue:44749", + "left": "issue:30990", + "right": "issue:31795", "accept": false, - "reason": "Compiled-file execution errors are unrelated to the post-upgrade slowdown report." + "reason": "Sentence-Transformer loading stall is not the same as confusing forward docs." }, { "left": "issue:43874", "right": "issue:46617", "accept": false, - "reason": "Missing image-processor method vs cache-loading FileNotFoundError; different components." + "reason": "Missing image-patch helper in GLM46V is unrelated to trust_remote_code cache loading." }, { - "left": "issue:30990", - "right": "issue:31795", + "left": "issue:43010", + "right": "issue:43232", "accept": false, - "reason": "Loading hang for Sentence Transformers is not the same as confusing docs for `input_ids`/`past_key_values`." + "reason": "Both involve generation internals, but they describe different failures and code paths." }, { "left": "issue:45850", "right": "issue:47110", "accept": false, - "reason": "A vague 'many models break' complaint is not the same as the specific `special_tokens` dict/list AttributeError." + "reason": "Broad v5 regressions vs a special_tokens dict/list attribute error; not the same underlying bug." }, { - "left": "issue:43010", - "right": "issue:43232", + "left": "issue:44230", + "right": "issue:46095", "accept": false, - "reason": "`update()` needing `@torch.no_grad` is unrelated to generation kwargs after `sync_gpus`." + "reason": "FP8 inference support request and weight-conversion question are different feature topics." }, { - "left": "issue:43502", - "right": "issue:44222", + "left": "issue:43299", + "right": "issue:44556", "accept": false, - "reason": "Network requests despite `local_files_only` is unrelated to FP8 moe save_pretrained." + "reason": "Qwen3VL-Moe loading breakage and old-checkpoint reload regression overlap broadly but are distinct reports." }, { - "left": "issue:43299", - "right": "issue:44556", + "left": "issue:38175", + "right": "issue:43502", + "accept": false, + "reason": "Zero probabilities in SigLIP2 is unrelated to unexpected API calls with local_files_only." + }, + { + "left": "issue:43827", + "right": "issue:44961", "accept": false, - "reason": "Qwen3VL MoE loading breakage is a different model-specific regression than generic checkpoint reload failure." + "reason": "Docs references to removed pipeline APIs are unrelated to the 'racoon' issue." }, { "left": "issue:46664", "right": "issue:46726", "accept": false, - "reason": "Video `num_frames` semantics and AutoProcessor vision-backend handling are different bugs." + "reason": "Video num_frames semantics and missing vision backend handling are different processor issues." }, { "left": "issue:30333", "right": "issue:36296", "accept": false, - "reason": "MLflow reporting on failed jobs is unrelated to tensor-parallel training failure." - }, - { - "left": "issue:43827", - "right": "issue:44961", - "accept": false, - "reason": "Docs still referencing `pipeline()` vs an unrelated 'racoon' issue; no shared bug." + "reason": "MLFlow reporting on failed jobs is unrelated to tensor parallel training." }, { "left": "issue:29870", "right": "issue:31795", "accept": false, - "reason": "Bart generate UnboundLocalError is unrelated to docs confusion about `input_ids` and `past_key_values`." + "reason": "BART generate() runtime error is unrelated to forward-argument documentation confusion." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "Qwen3-VL-Moe load bug vs missing `_tp_plan` for tensor parallelism; adjacent model family, different defect." - }, - { - "left": "issue:44230", - "right": "issue:46095", - "accept": false, - "reason": "FP8 inference support request is not the same as asking for a weight-conversion script." + "reason": "Both are Qwen3-Moe related, but one is a broad v5 bug and the other is a missing tensor-parallel plan." }, { "left": "issue:43576", "right": "issue:43824", "accept": false, - "reason": "Broken env command and a Qwen2.5-VL import error are different subsystems and failures." - }, - { - "left": "issue:43502", - "right": "issue:44961", - "accept": false, - "reason": "`local_files_only` network leakage is unrelated to the 'racoon' issue." + "reason": "CLI env command breakage and Qwen VL class import failure are unrelated." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled Python file errors are unrelated to PretrainedConfig type-checking regressions." - }, - { - "left": "issue:43576", - "right": "issue:45071", - "accept": false, - "reason": "CLI env command failure vs type-checking regression are not the same bug." + "reason": "Compiled-file execution errors are unrelated to PretrainedConfig type checking." }, { - "left": "issue:43576", - "right": "issue:47059", + "left": "issue:43502", + "right": "issue:44961", "accept": false, - "reason": "`transformers env` breakage is unrelated to CLI tests failing with `huggingface_hub` 1.22.0." + "reason": "Network access despite local_files_only and 'racoon' are unrelated." }, { - "left": "issue:42757", - "right": "issue:43576", + "left": "issue:43576", + "right": "issue:47059", "accept": false, - "reason": "Missing `is_offline_mode` import is a different import regression than the env command bug." + "reason": "transformers env CLI breakage and huggingface_hub completion attribute error are different dependency issues." }, { "left": "issue:36296", - "right": "issue:43901", + "right": "issue:43526", "accept": false, - "reason": "Tensor-parallel training bug is unrelated to docs mentioning deprecated `return_all_scores`." + "reason": "Tensor parallel training bug and BeitImageProcessorFast reduce_labels bug are unrelated." }, { "left": "issue:31795", "right": "issue:38617", "accept": false, - "reason": "Forward-method docs confusion is unrelated to missing `layer_type_validation` import." + "reason": "Docs confusion and layer_type_validation import error do not share the same bug." }, { - "left": "issue:43475", - "right": "issue:46617", + "left": "issue:42757", + "right": "issue:43576", "accept": false, - "reason": "SAM 3 video encoder attribute error vs trust_remote_code cache loading failure; different codepaths." + "reason": "Missing is_offline_mode import from huggingface_hub is unrelated to the env CLI command issue." }, { - "left": "issue:44077", - "right": "issue:44861", + "left": "issue:36683", + "right": "issue:47110", "accept": false, - "reason": "`post_init` policy validation is unrelated to `_get_tied_weight_keys` crashing on list-valued keys." + "reason": "Gemma3Config missing vocab_size and special_tokens keys error are different attribute problems." }, { - "left": "issue:44351", + "left": "issue:43475", "right": "issue:46617", "accept": false, - "reason": "`HybridCache` import failure is unrelated to trust_remote_code cache FileNotFoundError." + "reason": "SAM 3 Video missing fpn_position_embeddings is unrelated to trust_remote_code cache loading." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init acceptance and tied-weights list handling are separate configuration/utility bugs." } ] }, @@ -19910,7 +19984,7 @@ "nodes": [ "issue:41669", "issue:42371", - "issue:42890", + "issue:43298", "issue:43352", "issue:43408", "issue:43525", @@ -19923,6 +19997,7 @@ "issue:44295", "issue:44297", "issue:44315", + "issue:44351", "issue:44360", "issue:44393", "issue:44466", @@ -19931,15 +20006,12 @@ "issue:44556", "issue:44623", "issue:44655", - "issue:44704", "issue:44861", - "issue:44908", "issue:44961", "issue:45003", "issue:45071", "issue:45092", "issue:45230", - "issue:45341", "issue:45362", "issue:45375", "issue:45381", @@ -19948,324 +20020,312 @@ "issue:45753", "issue:45850", "issue:46018", + "issue:46095", "issue:46172", "issue:46209", "issue:46218", "issue:46328", + "issue:46617", "issue:46664", "issue:47059", "issue:47110", "issue:47455" ], "soft_pairs": [ - "issue:44623|issue:45230", "issue:45362|issue:45850", + "issue:44351|issue:46617", + "issue:44623|issue:45230", "issue:44315|issue:44360", "issue:45375|issue:47110", "issue:43673|issue:44492", + "issue:45381|issue:46664", + "issue:44273|issue:45542", "issue:44485|issue:44961", - "issue:43352|issue:44961", "issue:43525|issue:47110", - "issue:44273|issue:45542", - "issue:44961|issue:45468", - "issue:45381|issue:46664", + "issue:43352|issue:44961", "issue:44315|issue:46209", + "issue:44961|issue:45468", "issue:44273|issue:45003", - "issue:43867|issue:47110", "issue:44466|issue:44556", + "issue:43867|issue:47110", "issue:41669|issue:43673", - "issue:44655|issue:44861", + "issue:46018|issue:46172", "issue:43572|issue:45850", "issue:43525|issue:45850", - "issue:44485|issue:46664", + "issue:44655|issue:44861", "issue:44393|issue:45753", + "issue:44485|issue:46664", "issue:44393|issue:45071", - "issue:44295|issue:45230", "issue:44556|issue:45092", "issue:42371|issue:44492", + "issue:44295|issue:45230", "issue:44297|issue:45230", - "issue:45850|issue:47059", "issue:44246|issue:44273", - "issue:46018|issue:46172", - "issue:41669|issue:47455", + "issue:45850|issue:47059", "issue:43408|issue:44315", - "issue:44492|issue:45850", - "issue:41669|issue:46218", - "issue:43976|issue:44393", + "issue:41669|issue:47455", + "issue:43298|issue:46095", "issue:44492|issue:46328", - "issue:42890|issue:44908", - "issue:44704|issue:44908", - "issue:44908|issue:45071", - "issue:41669|issue:45341" + "issue:43976|issue:44393", + "issue:44492|issue:45850", + "issue:41669|issue:46218" ], - "prepared_review_unit_hash": "c9bfd08a06d7b57646c628eb6ebd9a65c995ee31293843c983e1043985bf6930", + "prepared_review_unit_hash": "12b1c11621379257835602d79911be6c52abbbc52ffe35073f5833e74ba58947", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 39, - "serialized_chars": 38635, - "estimated_input_tokens": 9659, - "estimated_eval_tokens": 19574 + "soft_pair_count": 37, + "serialized_chars": 38047, + "estimated_input_tokens": 9512, + "estimated_eval_tokens": 19280 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a noisy mix of unrelated bugs; the only coherent duplicate subgroup is around lazy loading/import behavior. I\u2019d consolidate that subgroup around issue 44273.", - "confidence": 0.57, - "canonical_issue_reason": "Issue 44273 is the clearest description of the underlying lazy-loading/import failure and sits at the center of the only plausible duplicate cluster (slow imports and sys.modules access).", + "summary": "This cluster is mostly heterogeneous: many items are different model- or feature-specific bugs that only share a loose Transformers v5/loading theme. The only clear duplicate-like edge is the lazy-loading/import-time slowdown pair.", + "confidence": 0.43, + "canonical_issue_reason": "issue 44273 is the most concrete root-cause issue here: it directly describes lazy loading being broken, which plausibly subsumes the slower-import symptom in 44246. The other issues are mostly separate bugs.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44273 is the best representative of the cluster because it names the core behavior directly and is the most technically central of the related import/lazy-loading reports.", + "best_issue_reason": "issue 44273 is the best representative issue for the only strong duplicate subcluster, and it is specific enough to anchor a fix without being as generic as the broad v5 regression reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:44623", - "right": "issue:45230", + "left": "issue:45362", + "right": "issue:45850", "accept": false, - "reason": "Generic bug report is not enough evidence of the same processor.save_pretrained missing-files bug." + "reason": "Generic 'v5 breaks models' complaint versus a specific Qwen crash; same theme, not the same bug." }, { - "left": "issue:45362", - "right": "issue:45850", + "left": "issue:44351", + "right": "issue:46617", + "accept": false, + "reason": "Both are load/import failures, but one is missing HybridCache import and the other is cache FileNotFoundError; different code paths." + }, + { + "left": "issue:44623", + "right": "issue:45230", "accept": false, - "reason": "One is a specific Qwen3.5 chat crash; the other is a broad compatibility complaint, not the same bug." + "reason": "A concrete save_pretrained files bug versus a placeholder bug report with no matching failure mode." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Different subsystems and failures: model_init/Liger Kernel vs a DSA indexer/ReLU discussion." + "reason": "Unrelated issues in different subsystems; no shared concrete bug." }, { "left": "issue:45375", "right": "issue:47110", "accept": false, - "reason": "Different config/parsing failures with no clear shared code path." + "reason": "Different problems: strict field dropping in a config versus special_tokens being a list." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Both mention cache, but one is chunked_prefill cache absence and the other is a cache-strategy typo." + "reason": "Chunked prefill cache regression versus a cache-strategy typo; not the same underlying defect." }, { - "left": "issue:44485", - "right": "issue:44961", + "left": "issue:45381", + "right": "issue:46664", "accept": false, - "reason": "GLM-5 RoPE implementation and racoon are unrelated." + "reason": "Both concern video/vision processing, but they are different model-specific bugs with different symptoms." }, { - "left": "issue:43352", - "right": "issue:44961", + "left": "issue:44273", + "right": "issue:45542", "accept": false, - "reason": "Flash Attention support issue vs unrelated racoon bug." + "reason": "Lazy-loading/import performance issue versus a TensorFlow backend installation bug." }, { - "left": "issue:43525", - "right": "issue:47110", + "left": "issue:44485", + "right": "issue:44961", "accept": false, - "reason": "Missing pad_token_id and a special_tokens keys error are different failures." + "reason": "No meaningful overlap beyond being issue titles; unrelated bugs." }, { - "left": "issue:44273", - "right": "issue:45542", + "left": "issue:43525", + "right": "issue:47110", "accept": false, - "reason": "Lazy-loading/import behavior is unrelated to the TensorFlow backend installation error." + "reason": "Missing pad_token_id in a config is unrelated to a special_tokens dict/list error." }, { - "left": "issue:44961", - "right": "issue:45468", + "left": "issue:43352", + "right": "issue:44961", "accept": false, - "reason": "Unrelated model-specific issues." + "reason": "Flash-Attention support warning versus an unrelated issue title." }, { - "left": "issue:45381", - "right": "issue:46664", + "left": "issue:44315", + "right": "issue:46209", "accept": false, - "reason": "Both are video-related, but the model behavior and code paths differ too much." + "reason": "Liger kernel application bug versus FP8 inference on B200; different layers and failure modes." }, { - "left": "issue:44315", - "right": "issue:46209", + "left": "issue:44961", + "right": "issue:45468", "accept": false, - "reason": "Different model/runtime problems with no shared concrete bug." + "reason": "No common bug; one is a placeholder title, the other is Gemma audio positional encoding." }, { "left": "issue:44273", "right": "issue:45003", - "accept": true, - "reason": "Both point to the same lazy-loading/import machinery; unsafe sys.modules access is a plausible root cause of lazy loading failing." - }, - { - "left": "issue:43867", - "right": "issue:47110", "accept": false, - "reason": "State_dict loading error and special_tokens dict/list error are unrelated." + "reason": "Both touch import/loading behavior, but one is lazy loading performance and the other is unsafe sys.modules access." }, { "left": "issue:44466", "right": "issue:44556", "accept": false, - "reason": "Related to serialization/loading in broad terms, but the concrete failure modes and scopes differ." + "reason": "Related to serialization/loading generally, but one is tied-weight serialization inconsistency and the other is checkpoint reload incompatibility." + }, + { + "left": "issue:43867", + "right": "issue:47110", + "accept": false, + "reason": "State_dict sorting load error versus special_tokens keys error; unrelated." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Import-star slowdown and chunked_prefill cache missing are different bugs." + "reason": "Import-star slowdown and chunked_prefill cache issues are both regressions, but not the same concrete bug." }, { - "left": "issue:44655", - "right": "issue:44861", + "left": "issue:46018", + "right": "issue:46172", "accept": false, - "reason": "Both touch save_pretrained, but pipeline saving and tied-weight-key type handling are not the same issue." + "reason": "Both are DeepSeek-V4 related, but one is gating logic and the other is a tensor transpose/cast issue." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "Missing pad_token_idx after update is unrelated to the broad breakage complaint." + "reason": "A specific StableLm config regression is not the same as a broad 'v5 breaks many models' report." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "Specific pad_token_id attribute error is not the same as a generic model breakage report." + "reason": "Specific config field regression versus a generic platform-wide complaint." }, { - "left": "issue:44485", - "right": "issue:46664", + "left": "issue:44655", + "right": "issue:44861", "accept": false, - "reason": "RoPE implementation and video num_frames semantics are unrelated." + "reason": "Both involve save_pretrained internals, but one is pipeline saving and the other is tied-weight-key handling." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Same model family, but 2D bbox output issues and infinite looping are different concrete bugs." + "reason": "Both are Qwen vision issues, but the symptoms and likely fixes differ." }, { - "left": "issue:44393", - "right": "issue:45071", + "left": "issue:44485", + "right": "issue:46664", "accept": false, - "reason": "Vision model output bug vs a type-checking regression are unrelated." + "reason": "Completely different model families and bugs." }, { - "left": "issue:44295", - "right": "issue:45230", + "left": "issue:44393", + "right": "issue:45071", "accept": false, - "reason": "Specific position_ids buffer bug vs generic bug report." + "reason": "Model-specific bounding box bug versus a global PretrainedConfig type-checking regression." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both are v5 compatibility complaints, but they affect different checkpoints and likely different root causes." + "reason": "Both are checkpoint-loading regressions after v5, but the titles point to different failure modes and models; not clearly one bug." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 settings API and cache-strategy typo are unrelated." + "reason": "TF32 API guidance is unrelated to cache strategy typos." }, { - "left": "issue:44297", + "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "Tokenizer serialization mismatch vs generic bug report." + "reason": "Specific position_ids buffer bug versus a generic bug report." }, { - "left": "issue:45850", - "right": "issue:47059", + "left": "issue:44297", + "right": "issue:45230", "accept": false, - "reason": "Broad model breakage complaint and CLI completion incompatibility are unrelated." + "reason": "Tokenizer save_pretrained mismatch versus a generic bug report." }, { "left": "issue:44246", "right": "issue:44273", "accept": true, - "reason": "Slow import and broken lazy loading are the same underlying import-path problem; slow imports are a plausible symptom of lazy loading failing." - }, - { - "left": "issue:46018", - "right": "issue:46172", - "accept": false, - "reason": "DeepSeek-V4 gating and a tensor transpose issue are different bugs." + "reason": "These describe the same underlying lazy-loading regression: broken lazy loading manifests as intermittent slow imports." }, { - "left": "issue:41669", - "right": "issue:47455", + "left": "issue:45850", + "right": "issue:47059", "accept": false, - "reason": "Import-star performance issue and kernel version checking are unrelated." + "reason": "Broad v5 regression complaint versus a specific CLI compatibility failure with huggingface_hub 1.22.0." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "Model type mismatch warning and Liger Kernel/model_init behavior are unrelated." - }, - { - "left": "issue:44492", - "right": "issue:45850", - "accept": false, - "reason": "Cache-strategy typo is too specific to be the same as a broad breakage report." + "reason": "Different model warning versus a training-time Liger kernel issue." }, { "left": "issue:41669", - "right": "issue:46218", + "right": "issue:47455", "accept": false, - "reason": "Import-star slowdown is unrelated to a missing module import for Qwen3_5ForCausalLM." + "reason": "Import-star performance regression versus an unrelated kernel-version check issue." }, { - "left": "issue:43976", - "right": "issue:44393", + "left": "issue:43298", + "right": "issue:46095", "accept": false, - "reason": "Python version compatibility and vision output correctness are different problems." + "reason": "Continuous batching support for audio models is unrelated to a DeepSeek-V4 weight-conversion question." }, { "left": "issue:44492", "right": "issue:46328", "accept": false, - "reason": "Cache-strategy typo and numpy/list padding inefficiency are unrelated." - }, - { - "left": "issue:42890", - "right": "issue:44908", - "accept": false, - "reason": "Test seeding and scheduler kwargs handling are different issues." + "reason": "Cache-strategy typo versus SequenceFeatureExtractor padding performance; different code paths." }, { - "left": "issue:44704", - "right": "issue:44908", + "left": "issue:43976", + "right": "issue:44393", "accept": false, - "reason": "AutoProcessor kwargs passthrough and inverse_sqrt scheduler kwargs are unrelated." + "reason": "Python version compatibility regression versus a Qwen vision output bug." }, { - "left": "issue:44908", - "right": "issue:45071", + "left": "issue:44492", + "right": "issue:45850", "accept": false, - "reason": "Scheduler kwargs bug and PretrainedConfig type checking regression are unrelated." + "reason": "A typo in cache strategies is not the same as the broad 'v5 breaks too many models' complaint." }, { "left": "issue:41669", - "right": "issue:45341", + "right": "issue:46218", "accept": false, - "reason": "Import performance issue and a testing_utils bug are unrelated." + "reason": "Import-time slowdown from import * is unrelated to the Qwen3_5 import failure." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays fairly conservative. The lazy-loading/import cluster around 44273 is the strongest coherent subgroup, and the accepted soft pairs are plausibly supported by the issue titles alone. No soft-edge verdict overstates the evidence in a way that clearly requires rejection." + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats the cluster as mostly heterogeneous, identifies the 44246\u219444273 lazy-loading/import slowdown pair as the only clear duplicate-like edge, and does not overclaim the weaker similarities. The soft-edge rejections are generally well supported by the issue titles." }, "evaluator_used": true, "retried": false, @@ -20292,9 +20352,9 @@ "issue:37428", "issue:41669", "issue:42371", + "issue:42890", "issue:42913", "issue:43066", - "issue:43298", "issue:43352", "issue:43502", "issue:43519", @@ -20302,6 +20362,7 @@ "issue:43606", "issue:43618", "issue:43704", + "issue:43827", "issue:43976", "issue:44230", "issue:44261", @@ -20312,10 +20373,11 @@ "issue:44485", "issue:44492", "issue:44623", - "issue:44829", + "issue:44704", "issue:44861", "issue:44908", "issue:44998", + "issue:45071", "issue:45230", "issue:45250", "issue:45290", @@ -20324,7 +20386,6 @@ "issue:45468", "issue:45656", "issue:45850", - "issue:46095", "issue:46218", "issue:46655", "issue:46894", @@ -20335,234 +20396,242 @@ "issue:47455" ], "soft_pairs": [ - "issue:43576|issue:44908", - "issue:43576|issue:47176", + "issue:44704|issue:44908", + "issue:42890|issue:44908", + "issue:41669|issue:45341", + "issue:44908|issue:45071", "issue:43976|issue:46218", - "issue:43298|issue:46095", + "issue:43827|issue:44908", + "issue:43576|issue:47176", "issue:45230|issue:45656", + "issue:42913|issue:43066", "issue:43976|issue:47176", "issue:31795|issue:33453", "issue:44273|issue:47455", "issue:44315|issue:47455", "issue:43519|issue:46655", - "issue:44315|issue:44829", - "issue:44861|issue:47110", "issue:45468|issue:46894", - "issue:42913|issue:43066", + "issue:44861|issue:47110", + "issue:43976|issue:47177", "issue:44485|issue:47176", + "issue:31795|issue:37428", "issue:44360|issue:45468", - "issue:43976|issue:47177", "issue:43519|issue:44485", - "issue:31795|issue:37428", "issue:43576|issue:44273", "issue:44485|issue:47177", "issue:44861|issue:47059", "issue:43519|issue:44908", "issue:41669|issue:43704", - "issue:43576|issue:46218", "issue:43606|issue:45468", - "issue:44485|issue:45468", + "issue:43576|issue:46218", + "issue:28348|issue:45250", "issue:42371|issue:43704", + "issue:44485|issue:45468", "issue:43704|issue:44485", - "issue:45850|issue:47177", - "issue:45850|issue:47176", "issue:43352|issue:47177", + "issue:43352|issue:47176", + "issue:45850|issue:47177", "issue:43502|issue:43519", "issue:43519|issue:44623", "issue:44261|issue:45468", - "issue:28348|issue:45250", "issue:45850|issue:46218", "issue:43704|issue:45341", "issue:43352|issue:46218", "issue:44351|issue:46218", - "issue:44861|issue:45290", "issue:44230|issue:44623", + "issue:44861|issue:45290", "issue:28348|issue:43352", + "issue:44230|issue:45310", "issue:44908|issue:44998", "issue:44998|issue:47177", "issue:44998|issue:47176", - "issue:44230|issue:45310", "issue:44492|issue:44998", "issue:43618|issue:44998" ], - "prepared_review_unit_hash": "def377abe17026b7b514e89630f3f920b3ad25bd051c6620fb60c1ddd28f802b", + "prepared_review_unit_hash": "c7d2bb75b389c6f861fe9398f65462e2d10d44102952f1907005b715c5d67c27", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 49, - "serialized_chars": 40989, - "estimated_input_tokens": 10248, - "estimated_eval_tokens": 20752 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 51, + "serialized_chars": 42040, + "estimated_input_tokens": 10510, + "estimated_eval_tokens": 21276 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "No true duplicate cluster stands out here. The items are mostly unrelated issue reports, with only loose thematic overlap around v5/tokenizer/import regressions and several model-specific bugs. All soft-edge candidates should be rejected under strict duplicate triage.", - "confidence": 0.62, - "canonical_issue_reason": "issue:43576 is the best anchor if a single issue must represent the set: it is concrete, version-specific, and has enough discussion/inbound references to be a useful hub for nearby v5-related reports. That said, it is not a duplicate of the others.", + "summary": "This cluster is highly heterogeneous: most items are unrelated bug reports across tokenization, loading, docs, security, processor, and model-specific regressions. The soft pairs are mostly superficial topical matches, not true duplicate artifacts.", + "confidence": 0.93, + "canonical_issue_reason": "No true canonical issue: the set is too mixed, and the soft similarities look like keyword/topic overlap rather than the same underlying bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43576 is the most suitable representative issue for triage because it has the clearest actionable failure report and broad compatibility/regression framing. It is still only a representative anchor, not evidence that the set is one duplicate bug.", + "best_issue_reason": "issue:43502 is the clearest standalone bug report with a concrete, reproducible loading regression, but it is still not representative of the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:43576", + "left": "issue:44704", "right": "issue:44908", "accept": false, - "reason": "Both mention v5-era behavior, but one is a broken CLI/env command and the other is an LR scheduler kwarg bug; different code paths and fixes." + "reason": "Different code paths: cached_file kwargs passthrough vs scheduler kwargs handling." }, { - "left": "issue:43576", - "right": "issue:47176", + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Training/test flakiness from missing set_seed is unrelated to scheduler behavior." + }, + { + "left": "issue:41669", + "right": "issue:45341", "accept": false, - "reason": "Completely different security/command-injection issue; only superficial CLI wording overlap." + "reason": "Import-star cleanup and a testing_utils bug are unrelated changes." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug is not the same as PretrainedConfig type-checking regression." }, { "left": "issue:43976", "right": "issue:46218", "accept": false, - "reason": "One is a Python version compatibility report, the other is a missing module import for a specific model; not the same bug." + "reason": "Python version compatibility and missing Qwen3_5 import are different failures." }, { - "left": "issue:43298", - "right": "issue:46095", + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Docs still mentioning pipeline() is unrelated to scheduler kwargs handling." + }, + { + "left": "issue:43576", + "right": "issue:47176", "accept": false, - "reason": "Continuous batching for audio models and DeepSeek weight conversion are unrelated problems in different subsystems." + "reason": "CLI env command breakage and shell-operator bypass are different issues." }, { "left": "issue:45230", "right": "issue:45656", "accept": false, - "reason": "Generic bug report versus a specific double optimizer-step issue under deepspeed; no shared concrete failure." + "reason": "Generic bug report versus double optimizer step; no shared code path." + }, + { + "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." }, { "left": "issue:43976", - "right": "issue:47176", + "right": "issue:47177", "accept": false, - "reason": "Python version support regression and shell-operator allowlist bypass are unrelated." + "reason": "Python support mismatch is unrelated to tar/zip file-write security." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Documentation confusion about forward arguments versus tokenizer loading regression; same area only at a very broad level." + "reason": "Documentation confusion about forward args is not a tokenizer loading regression." }, { "left": "issue:44273", "right": "issue:47455", "accept": false, - "reason": "Lazy-loading behavior and kernel-version checking are unrelated." + "reason": "Lazy loading failure and kernel version checking are unrelated." }, { "left": "issue:44315", "right": "issue:47455", "accept": false, - "reason": "Model_init/Liger application bug is unrelated to kernel version validation." + "reason": "Liger Kernel application under model_init is unrelated to version validation." }, { "left": "issue:43519", "right": "issue:46655", "accept": false, - "reason": "Both are processor bugs, but they affect different processors and different calculations (timestamp vs padding)." - }, - { - "left": "issue:44315", - "right": "issue:44829", - "accept": false, - "reason": "Different model-training issues: Liger application in model_init vs degenerate training with flash_attention_3." + "reason": "Both are processor bugs, but timestamp math and video padding are different logic paths." }, { - "left": "issue:44861", - "right": "issue:47110", + "left": "issue:45468", + "right": "issue:46894", "accept": false, - "reason": "Both are attribute/type errors, but in different APIs and different failure modes." + "reason": "Gemma audio positional encoding and ASR loss shifting are unrelated model bugs." }, { - "left": "issue:45468", - "right": "issue:46894", + "left": "issue:44861", + "right": "issue:47110", "accept": false, - "reason": "Gemma audio positional encoding and CohereASR double-shift training-loss bug are different model-specific problems." + "reason": "Weight-key list crash is not the same as special_tokens list handling." }, { - "left": "issue:42913", - "right": "issue:43066", + "left": "issue:43976", + "right": "issue:47177", "accept": false, - "reason": "Both concern tokenizer v5 changes, but one is a broad behavior difference and the other is a specific decoder-type bug; not the same concrete defect." + "reason": "Python compatibility issue is unrelated to shell/file extraction exploits." }, { "left": "issue:44485", "right": "issue:47176", "accept": false, - "reason": "RoPE implementation discussion and shell command-substitution bypass are unrelated." + "reason": "RoPE implementation bug and command-substitution allowlist bypass are unrelated." }, { - "left": "issue:44360", - "right": "issue:45468", + "left": "issue:31795", + "right": "issue:37428", "accept": false, - "reason": "DSA ReLU discussion and Gemma audio positional encoding are unrelated model/implementation issues." + "reason": "Forward-argument docs confusion is unrelated to missing flash-support import symbol." }, { - "left": "issue:43976", - "right": "issue:47177", + "left": "issue:44360", + "right": "issue:45468", "accept": false, - "reason": "Python version compatibility and tar/zip slip security issue are unrelated." + "reason": "DSA ReLU omission is unrelated to Gemma audio positional encoding." }, { "left": "issue:43519", - "right": "issue:44908", - "accept": false, - "reason": "Different bugs in different components: Qwen3VL timestamp calculation vs scheduler kwarg handling." - }, - { - "left": "issue:31795", - "right": "issue:37428", + "right": "issue:44485", "accept": false, - "reason": "Documentation confusion versus a flash-attention import error; no shared code-path bug." + "reason": "Processor timestamp math and GLM-5 RoPE are different subsystems." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "A broken env command and lazy loading failure are different issues, even if both appear in v5 contexts." + "reason": "CLI env command failure does not match lazy-loading behavior." }, { "left": "issue:44485", "right": "issue:47177", "accept": false, - "reason": "RoPE implementation and arbitrary file write/RCE are entirely unrelated." + "reason": "Model RoPE math and shell injection/file-write security are unrelated." }, { "left": "issue:44861", "right": "issue:47059", "accept": false, - "reason": "Different attribute errors in different areas; no evidence of one underlying defect." + "reason": "Weight-key attribute crash is unrelated to Hugging Face CLI completion breakage." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Processor timestamp bug versus scheduler kwarg bug; different subsystem and fix." + "reason": "Processor timestamp calculation is unrelated to scheduler kwargs." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-star cleanup/performance complaint is unrelated to VRAM leakage in multithreaded dataloaders." - }, - { - "left": "issue:43576", - "right": "issue:46218", - "accept": false, - "reason": "CLI/env command breakage and missing model-module import are unrelated failure modes." + "reason": "Import-star removal is unrelated to a VRAM leak in Qwen3ForCausalLM." }, { "left": "issue:43606", @@ -20571,142 +20640,148 @@ "reason": "CPU offload device mismatch and Gemma audio positional encoding are unrelated." }, { - "left": "issue:44485", - "right": "issue:45468", + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command and missing model import are different failures." + }, + { + "left": "issue:28348", + "right": "issue:45250", "accept": false, - "reason": "RoPE implementation and audio positional encoding are not the same bug." + "reason": "Both mention Flash Attention 2.0, but one is GPT2-specific support and the other is too generic to treat as the same bug/change." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 configuration guidance and VRAM leakage are unrelated." + "reason": "TF32 API guidance and VRAM leakage are unrelated." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Different model math issues with no shared code path." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak in multi-threaded loading and GLM-5 RoPE implementation are unrelated." + "reason": "VRAM leak in multithreaded dataloaders is unrelated to RoPE implementation." }, { - "left": "issue:45850", + "left": "issue:43352", "right": "issue:47177", "accept": false, - "reason": "A broad v5 breakage complaint and a tar/zip slip RCE are unrelated." + "reason": "Flash-Attention support limitation and archive extraction RCE are unrelated." }, { - "left": "issue:45850", + "left": "issue:43352", "right": "issue:47176", "accept": false, - "reason": "Broad v5 breakage and command-substitution allowlist bypass are unrelated." + "reason": "Model Flash Attention support and shell-operator bypass are unrelated." }, { - "left": "issue:43352", + "left": "issue:45850", "right": "issue:47177", "accept": false, - "reason": "Flash Attention 2.0 support gap and archive extraction RCE are unrelated." + "reason": "Broad version-breakage complaint and file-write/RCE issue are not duplicates." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "Local-files-only API requests and Qwen3VL timestamp calculation are different bugs." + "reason": "Local_files_only network requests and processor timestamp math are different bugs." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Processor timestamp bug versus missing processor.save_pretrained files; different symptoms and fixes." + "reason": "Processor timestamp bug is unrelated to missing processor.save_pretrained files." }, { "left": "issue:44261", "right": "issue:45468", "accept": false, - "reason": "MLA rms_norm_eps precision issue and Gemma audio positional encoding are unrelated model implementation bugs." - }, - { - "left": "issue:28348", - "right": "issue:45250", - "accept": false, - "reason": "Specific GPT2LMHeadModel Flash Attention 2.0 support request and a vague Flash Attention 2.0 issue are not enough to prove the same concrete bug." + "reason": "RMS norm eps precision bug is unrelated to audio positional encoding." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "Broad v5 breakage complaint and model import failure are unrelated." + "reason": "General v5 breakage is too broad to merge with a specific missing import error." }, { "left": "issue:43704", "right": "issue:45341", "accept": false, - "reason": "VRAM leakage in dataloader threads and a testing_utils bug are unrelated." + "reason": "VRAM leak in training threads and a testing_utils bug are unrelated." }, { "left": "issue:43352", "right": "issue:46218", "accept": false, - "reason": "Flash Attention 2.0 support gap for Nemotron and Qwen module import failure are different issues." + "reason": "Flash Attention support for NemotronH is unrelated to missing Qwen3_5 import." }, { "left": "issue:44351", "right": "issue:46218", "accept": false, - "reason": "Missing HybridCache import and missing Qwen3_5ForCausalLM module are both import-related, but they are not the same broken symbol or code path." + "reason": "HybridCache import failure and Qwen3_5 import failure are different symbols and code paths." }, { - "left": "issue:44861", - "right": "issue:45290", + "left": "issue:44230", + "right": "issue:44623", "accept": false, - "reason": "Tied-weight-key attribute crash and chat-template tool-call crash are unrelated." + "reason": "FP8 model support and missing processor save files are unrelated." }, { - "left": "issue:44230", - "right": "issue:44623", + "left": "issue:44861", + "right": "issue:45290", "accept": false, - "reason": "FP8 inference support request and processor.save_pretrained missing files are unrelated." + "reason": "Weight-key crash is unrelated to chat template tool-call handling." }, { "left": "issue:28348", "right": "issue:43352", "accept": false, - "reason": "Both mention Flash Attention 2.0, but one is GPT2 support and the other is Nemotron support; different models and concrete changes." + "reason": "Both are Flash Attention 2.0 related, but one is GPT2 support and the other is Nemotron-specific lack of support." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 support request and Qwen3.5 MoE from_pretrained error are not the same bug." }, { "left": "issue:44908", "right": "issue:44998", "accept": false, - "reason": "Inverse-sqrt scheduler kwargs bug and an unrelated non-technical issue clearly do not match." + "reason": "Scheduler kwargs bug and an unrelated placeholder issue are not duplicates." }, { "left": "issue:44998", "right": "issue:47177", "accept": false, - "reason": "Unemployment joke/noise issue and archive RCE are unrelated." + "reason": "Placeholder issue is unrelated to archive extraction security." }, { "left": "issue:44998", "right": "issue:47176", "accept": false, - "reason": "Noise issue and command-injection bypass are unrelated." - }, - { - "left": "issue:44230", - "right": "issue:45310", - "accept": false, - "reason": "FP8 support request and Qwen3.5 MoE from_pretrained error are related only at the model family level, not the same bug." + "reason": "Placeholder issue is unrelated to command-substitution bypass." }, { "left": "issue:44492", "right": "issue:44998", "accept": false, - "reason": "Typo in cache strategies and unrelated noise are not duplicates." + "reason": "Cache-strategy typo and unrelated placeholder issue are not duplicates." }, { "left": "issue:43618", "right": "issue:44998", "accept": false, - "reason": "CLIPOutput attentions regression and unrelated noise are not duplicates." + "reason": "CLIP attentions regression is unrelated to the placeholder issue." } ] }, @@ -20750,8 +20825,8 @@ ], "soft_pairs": [ "issue:31515|issue:38617", - "issue:29870|issue:31515", "issue:43298|issue:44230", + "issue:29870|issue:31515", "issue:30333|issue:31515", "issue:31515|issue:36296", "issue:31515|issue:31795", @@ -20768,142 +20843,142 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "0873cd7169c3b8323813dffbb9d04af4056b49c29f4cc6aebb72e7fbeffb2063", + "prepared_review_unit_hash": "0cf3db67b7beda680a3d6b2e63ebad817ee96ff52d4df6e1df30801d06b9ed41", "component_budget": { "node_count": 544, "item_count": 544, - "soft_pair_count": 2017, - "serialized_chars": 1043917, - "estimated_input_tokens": 260980, - "estimated_eval_tokens": 522216 + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 }, "budget": { "node_count": 17, "item_count": 17, "soft_pair_count": 18, - "serialized_chars": 14983, - "estimated_input_tokens": 3746, - "estimated_eval_tokens": 7748 + "serialized_chars": 14985, + "estimated_input_tokens": 3747, + "estimated_eval_tokens": 7750 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous; the high similarity scores are mostly driven by generic model/config wording rather than a shared root cause. None of the soft-edge pairs look like the same underlying bug or change, so no merges are justified.", - "confidence": 0.96, - "canonical_issue_reason": "No single canonical issue: the items span unrelated topics such as generation failures, training/reporting, loading speed, docs, tensor parallelism, config typing, and model-specific implementation bugs.", - "canonical_pr_reason": "No pull requests are present in this cluster.", - "best_issue_reason": "None of the issues is a good global representative. Issue 31515 is only a superficial hub because it mentions generic checkpoint/loading language, but it does not share a concrete root cause with the others.", - "best_pr_reason": "No pull requests are present in this cluster.", + "summary": "All proposed soft edges look like false positives: the issues span unrelated subsystems and symptoms (loading, training, docs, config typing, model behavior, tokenizer warnings, etc.). None of the pairs appear to describe the same underlying bug or change.", + "confidence": 0.98, + "canonical_issue_reason": "No clear canonical issue stands out because the cluster is heterogeneous; issue:31515 is only the most connected by soft similarity, not a semantic duplicate target for the others.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue for consolidation: none of the items are true duplicates, so selecting any single issue would misrepresent the cluster.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:31515", "right": "issue:38617", "accept": false, - "reason": "Checkpoint loading slowness vs missing import error; unrelated problems and code paths." + "reason": "Different problems: checkpoint loading performance vs an import error in configuration utilities." }, { - "left": "issue:29870", - "right": "issue:31515", + "left": "issue:43298", + "right": "issue:44230", "accept": false, - "reason": "BART generate decoder mask bug vs slow checkpoint loading; different symptoms and subsystems." + "reason": "Unrelated topics: continuous batching audio support vs fp8 inference support for Qwen VL/MoE models." }, { - "left": "issue:43298", - "right": "issue:44230", + "left": "issue:29870", + "right": "issue:31515", "accept": false, - "reason": "Continuous batching audio-model support vs fp8 Qwen VL/MoE inference support; different model features." + "reason": "BART generate() UnboundLocalError is unrelated to slow from_pretrained checkpoint loading." }, { "left": "issue:30333", "right": "issue:31515", "accept": false, - "reason": "MLflow job-state reporting vs checkpoint loading performance; no shared bug." + "reason": "MLFlow reporting on failed training jobs is a separate tracking/reporting issue, not checkpoint loading." }, { "left": "issue:31515", "right": "issue:36296", "accept": false, - "reason": "Loading speed issue vs tensor-parallel training bug; unrelated." + "reason": "Checkpoint loading speed and tensor-parallel training bugs are different code paths and failures." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Loading performance vs confusing forward-docs issue; one is a runtime bug, the other documentation." + "reason": "Documentation confusion about forward arguments is not the same as slow checkpoint loading." }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba init-weights behavior vs RoFormer attention interface usage; different models and fixes." + "reason": "Mamba initialization behavior and RoFormer attention-interface usage are different model-specific implementation issues." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former ignore_value handling vs DeepSeek-V4 shared-expert gating; unrelated model internals." + "reason": "Mask2Former ignore_value handling and DeepSeek-V4 shared-expert gating are unrelated model bugs." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "ModernBERT type-hint issue vs tokenizer regex warning regression; unrelated." + "reason": "A config type-hint typo is unrelated to tokenizer regex warning behavior." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Ignored value not used vs kwargs not passed to loss calculation; different bug locations." + "reason": "Missing ignore_value usage in Mask2Former is not the same as kwargs propagation to a loss function." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss kwargs propagation vs RoFormer attention interface usage; no same code path." + "reason": "Loss kwargs plumbing and RoFormer attention interface usage affect different parts of the stack." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former ignore_value bug vs tokenizer warning regression; unrelated." + "reason": "Mask2Former loss-value handling is unrelated to tokenizer regex warnings." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Mask2Former loss-argument handling vs RoFormer attention interface usage; unrelated." + "reason": "These concern different models and bugs: Mask2Former ignore_value vs RoFormer attention API usage." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Mask2Former ignore_value bug vs typo in quantization operation text; not the same issue." + "reason": "Mask2Former loss handling and a quantization-operation typo are different issues." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Incorrect type hint in config vs quantization-operation typo; unrelated." + "reason": "Incorrect type hint in ModernBERTConfig is unrelated to a typo in quantization conversion code." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwargs propagation vs ModernBERT config typing; different components." + "reason": "Passing kwargs to loss calculation is unrelated to a config annotation problem." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Llama4TextL2Norm/RMSNorm question vs ModernBERT type hint issue; unrelated." + "reason": "A question about Llama4TextL2Norm behavior is not the same as a ModernBERT config type-hint bug." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value bug vs ModernBERT type hint issue; no common underlying change." + "reason": "Mask2Former ignore_value usage and ModernBERT norm_eps typing are unrelated." } ] }, @@ -20936,7 +21011,7 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "122fb8d5f6d296369fd24d321ca83a5840aa92481ea350037842fd81d7307786", + "prepared_review_unit_hash": "6bf874f3c6793fe26c857b2ac4fbc84dcfc4821dbf3d416c871fc07a189840c0", "component_budget": { "node_count": 3, "item_count": 3, @@ -20958,30 +21033,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three unrelated feature requests: tokenizer truncation direction, ReduceLROnPlateau-like scheduler behavior, and Wav2Vec2 word-level timing offsets. They do not describe the same underlying bug/change, so no duplicate cluster emerges.", + "summary": "The cluster is not a duplicate set: it mixes three unrelated feature requests in tokenization, learning-rate scheduling, and Wav2Vec2 alignment. No pair looks like the same underlying issue.", "confidence": 0.98, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:11307 has the strongest discussion activity, but it is still unrelated to the other two issues, so there is no true canonical issue for this cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different areas and different requested capabilities: truncation strategy vs Wav2Vec2 word offsets." + "reason": "10519 is about truncating from the beginning in tokenization; 11307 is about getting word start/end offsets in Wav2Vec2. Different components and different problems." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "No shared bug or code path; one is about tokenizer truncation direction, the other about learning-rate scheduler behavior." + "reason": "10519 requests a truncation direction option, while 11005 asks for ReduceLROnPlateau-like scheduler behavior. No shared code path or underlying bug." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "Unrelated functionality requests in different subsystems; no plausible single fix would cover both." + "reason": "11005 concerns training scheduler functionality; 11307 concerns speech-model time offsets. These are unrelated feature requests." } ] }, @@ -21013,12 +21088,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "90ab2797731d3f54af06011c8067bf3aec30af600d024d3874cf4f20481927ee", + "prepared_review_unit_hash": "ee4e442dc388044b6a830d55d6b5144b359990322fd242076dde0bdc0ed2ffd8", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2479, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -21026,7 +21101,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2479, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -21035,24 +21110,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only issues are present here. `issue:16618` is the best cluster anchor because it is the broadest model-loading complaint and the only one with any plausible overlap with another item; `issue:15980` is a narrower auth/download error, and `issue:13244` is unrelated tokenization behavior.", - "confidence": 0.88, - "canonical_issue_reason": "`issue:16618` is the most central of the three: it concerns failing to load a model, which is the broadest and most likely umbrella symptom in this set. The others are either a specific login/download error (`15980`) or a separate tokenization discrepancy (`13244`).", + "summary": "These issues do not appear to describe the same underlying bug: one is about Tapas tokenization behavior, one about an authentication/error-message problem when downloading private models, and one about loading bert-base-uncased. The soft links look like broad topical similarity only.", + "confidence": 0.91, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "`issue:16618` is the strongest representative issue because it matches the cluster\u2019s shared theme of model-loading failure more than the others and is the most general user-facing problem statement.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Different problems: tokenization mismatch vs. model-loading failure. They do not describe the same underlying bug." + "reason": "Tapas tokenization mismatch and generic model loading failure are different code-path problems; no clear shared root cause." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "Related only at a very high level. One is a bad login/download error message for private models; the other is a generic inability to load `bert-base-uncased`. Not clearly the same bug." + "reason": "Private-model login/download error messaging is distinct from a general inability to load bert-base-uncased." } ] }, @@ -21074,7 +21149,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-19631-8", + "cluster_id": "cluster-19631-10", "nodes": [ "issue:19631", "issue:19865", @@ -21082,8 +21157,10 @@ "issue:20362", "issue:22352", "issue:22372", + "issue:23050", "issue:25725", - "issue:25839" + "issue:25839", + "issue:28306" ], "soft_pairs": [ "issue:19631|issue:22372", @@ -21096,102 +21173,116 @@ "issue:20362|issue:25839", "issue:20362|issue:25725", "issue:22352|issue:25725", - "issue:22372|issue:25725" + "issue:22372|issue:25725", + "issue:23050|issue:28306", + "issue:23050|issue:25725" ], - "prepared_review_unit_hash": "d02f83e3fa0702c6ea56f3e7d1efa22e7d1986b66767de31ddb1d8d3dbe6d3a1", + "prepared_review_unit_hash": "41ef4761d7e29dfafe647657336b00a125e19dff5d21d292d2537080f973b768", "component_budget": { - "node_count": 8, - "item_count": 8, - "soft_pair_count": 11, - "serialized_chars": 8196, - "estimated_input_tokens": 2049, - "estimated_eval_tokens": 4354 + "node_count": 10, + "item_count": 10, + "soft_pair_count": 13, + "serialized_chars": 9966, + "estimated_input_tokens": 2492, + "estimated_eval_tokens": 5240 }, "budget": { - "node_count": 8, - "item_count": 8, - "soft_pair_count": 11, - "serialized_chars": 8196, - "estimated_input_tokens": 2049, - "estimated_eval_tokens": 4354 + "node_count": 10, + "item_count": 10, + "soft_pair_count": 13, + "serialized_chars": 9966, + "estimated_input_tokens": 2492, + "estimated_eval_tokens": 5240 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates; they are separate feature requests for different models or model-related tasks. The shared wording is just a common \"Add model\" pattern, not the same underlying change.", + "summary": "These issues are all distinct feature requests for different models or model-related tasks. The soft similarities come from being in the same broad \u2018add a model\u2019 area, but there is no evidence they describe the same underlying change or bug.", "confidence": 0.96, - "canonical_issue_reason": null, + "canonical_issue_reason": "No true canonical duplicate: the cluster is heterogeneous, covering unrelated model additions (EDSR/MDSR, VATT, FlexiBERT, Restormer, LLaMA-MoE, etc.). If forced to pick the broadest issue, issue:25725 (Meta-Transformer) is the most generic, but it does not subsume the others.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:25725 is the broadest/most generic model-request title in the set, but it is still not a duplicate representative for the rest; the items are separate requests for different architectures and tasks.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:19631", "right": "issue:22372", "accept": false, - "reason": "EDSR/MDSR vs Restormer are different model additions with no concrete overlap." + "reason": "EDSR/MDSR super-resolution and Restormer are different model families and different requested additions." }, { "left": "issue:19631", "right": "issue:20362", "accept": false, - "reason": "EDSR/MDSR and FlexiBERT are unrelated model requests." + "reason": "EDSR/MDSR and FlexiBERT are unrelated model requests with no shared code-path or concrete duplicate target." }, { "left": "issue:19865", "right": "issue:25725", "accept": false, - "reason": "VATT and Meta-Transformer are distinct model targets." + "reason": "VATT and Meta-Transformer are different research models; similarity is only that both are model-addition requests." }, { "left": "issue:19865", "right": "issue:22372", "accept": false, - "reason": "VATT and Restormer are different requested additions." + "reason": "VATT (audio/vision transformer) and Restormer (image restoration) are distinct changes." }, { "left": "issue:20334", "right": "issue:22352", "accept": false, - "reason": "A longformer content issue and Whisper XVector finetuning are not the same change." + "reason": "longformer_content and Whisper XVector are unrelated tasks and model integrations." }, { "left": "issue:20334", "right": "issue:20362", "accept": false, - "reason": "Longformer-related content and FlexiBERT are different model requests." + "reason": "longformer_content and FlexiBERT are different model requests, not the same underlying change." }, { "left": "issue:20334", "right": "issue:25725", "accept": false, - "reason": "Longformer content and Meta-Transformer are unrelated feature requests." + "reason": "longformer_content and Meta-Transformer refer to different models and use cases." }, { "left": "issue:20362", "right": "issue:25839", "accept": false, - "reason": "FlexiBERT and D_Nikud are separate model additions." + "reason": "FlexiBERT and D_Nikud are separate model additions for different NLP tasks." }, { "left": "issue:20362", "right": "issue:25725", "accept": false, - "reason": "FlexiBERT and Meta-Transformer are different concrete model requests." + "reason": "FlexiBERT and Meta-Transformer are different architectures and request contexts." }, { "left": "issue:22352", "right": "issue:25725", "accept": false, - "reason": "Whisper XVector finetuning and Meta-Transformer do not describe the same bug or change." + "reason": "Whisper XVector finetuning and Meta-Transformer are unrelated; they do not look mergeable into one fix." }, { "left": "issue:22372", "right": "issue:25725", "accept": false, - "reason": "Restormer and Meta-Transformer are distinct model additions." + "reason": "Restormer and Meta-Transformer are different models serving different purposes." + }, + { + "left": "issue:23050", + "right": "issue:28306", + "accept": false, + "reason": "TTS advanced and LLaMA-MoE are separate model requests in different domains." + }, + { + "left": "issue:23050", + "right": "issue:25725", + "accept": false, + "reason": "Advanced TTS and Meta-Transformer are not the same underlying feature or code path." } ] }, @@ -21289,12 +21380,12 @@ "issue:47299|issue:47402", "issue:46226|issue:47299" ], - "prepared_review_unit_hash": "80ef1bd0cb6bc06571a781bc6c39bb47ae1f435b8f0b74ae3165d9e92e05b3ee", + "prepared_review_unit_hash": "3a0a7a40003b31b451a4c64578697f1c10bcb8ae7ea48dfde5f09d44a6129d5b", "component_budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 55, - "serialized_chars": 29305, + "serialized_chars": 29306, "estimated_input_tokens": 7327, "estimated_eval_tokens": 14910 }, @@ -21302,7 +21393,7 @@ "node_count": 16, "item_count": 16, "soft_pair_count": 55, - "serialized_chars": 29305, + "serialized_chars": 29306, "estimated_input_tokens": 7327, "estimated_eval_tokens": 14910 }, @@ -21311,352 +21402,349 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is mostly unrelated i18n translation requests. The only plausible duplicates are the exact placeholder-title translation issues, which appear to be the same template-based request repeated multiple times; the language-specific translation requests are distinct and should stay separate. The Flash Attention 2 and GGUF support issues are different feature requests.", - "confidence": 0.77, - "canonical_issue_reason": "Issue 43834 is the cleanest canonical issue for the duplicate placeholder-title group: it is the earliest of the near-identical template translation requests and best represents that repeated issue pattern.", + "summary": "This cluster is mostly a set of distinct i18n translation-request issues with highly similar template wording, plus two unrelated feature requests (Flash Attention 2 and GGUF support). None of the soft edges look like true duplicates of the same underlying change.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43834 is the cleanest generic i18n template and the best representative of the repeated translation-request family, but the neighboring issues still appear to be separate language-specific requests rather than duplicates.", "canonical_pr_reason": null, - "best_issue_reason": "43834 is the best issue to keep as the representative anchor because it is the earliest and most central of the exact-match template translation issues.", + "best_issue_reason": "issue:43834 is the most central and most broadly similar issue in the translation subgroup, so it is the best single issue to anchor the cluster even though it should not absorb the others as duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Both are broad community contribution feature requests, but they target different capabilities (Flash Attention 2 vs GGUF) and are not the same concrete change." + "reason": "Different feature areas: Flash Attention 2 support vs GGUF support. Same broad contribution style, but not the same underlying change." }, { "left": "issue:26794", "right": "issue:44370", "accept": false, - "reason": "Different translation requests; same i18n template style is not enough to make them duplicates." + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." }, { "left": "issue:26794", "right": "issue:44365", "accept": false, - "reason": "Different languages/request targets, so these are separate issues." + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." }, { "left": "issue:26794", "right": "issue:47402", "accept": false, - "reason": "Same generic translation template, but no evidence they are the same underlying language request." + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." }, { "left": "issue:26794", "right": "issue:43834", "accept": false, - "reason": "Different language-specific translation requests; not the same issue." + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." }, { "left": "issue:26794", "right": "issue:47299", "accept": false, - "reason": "Distinct translation request, not a duplicate of the Kannada issue." + "reason": "The titles do not indicate the same language or request; no evidence these are the same translation task." }, { "left": "issue:26794", "right": "issue:27088", "accept": false, - "reason": "Kannada and Turkish translation requests are different issues." + "reason": "Different target languages are implied; these are separate translation requests, not a duplicate." }, { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Different translation requests; shared template text is insufficient for duplication." + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different language translation requests, so not duplicates." + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." }, { "left": "issue:27088", "right": "issue:47402", "accept": false, - "reason": "No concrete evidence these are the same underlying translation issue." + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "Different languages/request targets; should not be merged." + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish translation request is distinct from the placeholder template issues." + "reason": "Spanish translation vs a generic i18n template issue; same subsystem, but not the same language request." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish and Spanish translation issues are different." + "reason": "Turkish and Spanish translation requests are separate language tasks." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Different language translation requests." + "reason": "Different translation requests; no evidence they refer to the same language or work item." }, { "left": "issue:28936", "right": "issue:30240", "accept": false, - "reason": "Spanish and Polish translation requests are separate issues." + "reason": "Spanish and Polish translation requests are separate language tasks." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Spanish and Romanian translation requests are different." + "reason": "Spanish and Romanian translation requests are separate language tasks." }, { "left": "issue:27088", "right": "issue:30191", "accept": false, - "reason": "Turkish and Persian translation requests are distinct." + "reason": "Turkish and Persian translation requests are separate language tasks." }, { "left": "issue:26794", "right": "issue:30191", "accept": false, - "reason": "Kannada and Persian translation requests are separate issues." + "reason": "Kannada and Persian translation requests are separate language tasks." }, { "left": "issue:30191", "right": "issue:30240", "accept": false, - "reason": "Persian and Polish translation requests are different." + "reason": "Persian and Polish translation requests are separate language tasks." }, { "left": "issue:28936", "right": "issue:30191", "accept": false, - "reason": "Spanish and Persian translation requests are not duplicates." + "reason": "Spanish and Persian translation requests are separate language tasks." }, { "left": "issue:26794", "right": "issue:30240", "accept": false, - "reason": "Kannada and Polish translation requests are different issues." + "reason": "Kannada and Polish translation requests are separate language tasks." }, { "left": "issue:30240", "right": "issue:44370", "accept": false, - "reason": "Polish translation request is distinct from the generic placeholder issues." + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:30240", "right": "issue:44365", "accept": false, - "reason": "Different translation requests; no evidence of the same underlying issue." + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:30240", "right": "issue:47402", "accept": false, - "reason": "Polish translation request is separate from the placeholder template issues." + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:30240", "right": "issue:43834", "accept": false, - "reason": "Different underlying translation requests." + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian translation request is distinct from the placeholder template issues." + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Different translation requests." + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:38435", "right": "issue:47402", "accept": false, - "reason": "Romanian translation issue is not the same as the generic placeholder issue." + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "Different translation requests; same subsystem is too broad to merge." + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." }, { "left": "issue:38435", "right": "issue:47299", "accept": false, - "reason": "Romanian and the Thai-text issue are not clearly the same underlying request." + "reason": "Different translation request content; no sign these are the same underlying issue." }, { "left": "issue:43834", "right": "issue:44370", - "accept": true, - "reason": "Exact same placeholder translation title and near-identical issue body indicate the same underlying template issue." + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." }, { "left": "issue:43834", "right": "issue:44365", - "accept": true, - "reason": "These are effectively identical template translation issues with no meaningful distinction visible." + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." }, { "left": "issue:43834", "right": "issue:47402", - "accept": true, - "reason": "Same placeholder title/body pattern; looks like a duplicate of the same translation template issue." + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." }, { "left": "issue:43834", "right": "issue:47299", "accept": false, - "reason": "Despite high similarity, the titles/body text differ and there is no clear evidence it is the same language request." + "reason": "Template similarity alone is not enough; the underlying translation request is not shown to be the same." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Different visible title content; not enough evidence for a duplicate merge." - }, - { - "left": "issue:43834", - "right": "issue:46931", - "accept": false, - "reason": "The title text is different and no concrete same-underlying-issue evidence is present." - }, - { - "left": "issue:43834", - "right": "issue:46222", - "accept": false, - "reason": "Different visible title content; could be similar spammy translation requests, but not the same issue." + "reason": "Different issue text and no evidence of the same translation request." }, { "left": "issue:44365", "right": "issue:44370", - "accept": true, - "reason": "Same placeholder translation issue pattern; plausible duplicate." + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." }, { "left": "issue:44365", "right": "issue:47402", - "accept": true, - "reason": "Same generic template translation request with effectively identical content." + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." }, { "left": "issue:44365", "right": "issue:47299", "accept": false, - "reason": "Different visible content; not safe to treat as the same issue." + "reason": "Identical template wording does not prove the same language request; treat as separate issues." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Titles differ and the bodies do not clearly show the same concrete request." - }, - { - "left": "issue:44365", - "right": "issue:46931", - "accept": false, - "reason": "Different visible issue content; no solid duplicate signal." - }, - { - "left": "issue:44365", - "right": "issue:46222", - "accept": false, - "reason": "Similar theme, but not the same underlying issue." + "reason": "Different issue text and no evidence they are the same translation task." }, { "left": "issue:44370", "right": "issue:47402", - "accept": true, - "reason": "Exact same placeholder translation request pattern; acceptable duplicate." + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." }, { "left": "issue:44370", "right": "issue:47299", "accept": false, - "reason": "Different visible title/body content, so not clearly the same issue." + "reason": "Identical template wording does not prove the same language request; treat as separate issues." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Not enough evidence these are the same underlying translation request." + "reason": "Different issue text and no evidence they are the same translation task." }, { - "left": "issue:44370", - "right": "issue:46931", + "left": "issue:46222", + "right": "issue:47299", "accept": false, - "reason": "Different visible content; should remain separate." + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they represent the same request." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Similar theme only, not the same concrete issue." + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." }, { - "left": "issue:46222", - "right": "issue:47299", + "left": "issue:44365", + "right": "issue:46222", "accept": false, - "reason": "Both are Thai-text issues, but the visible wording differs and there is no clear duplicate match." + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." }, { "left": "issue:46222", "right": "issue:47402", "accept": false, - "reason": "Different visible content; not the same issue." + "reason": "No evidence these are the same underlying translation request." }, { - "left": "issue:46226", - "right": "issue:46931", + "left": "issue:43834", + "right": "issue:46222", "accept": false, - "reason": "Different Thai-text issue titles; no concrete duplicate evidence." + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." }, { "left": "issue:46226", - "right": "issue:47299", + "right": "issue:47402", "accept": false, - "reason": "Similar language/script is not enough to merge these as duplicates." + "reason": "No evidence these are the same underlying translation request." }, { "left": "issue:46226", - "right": "issue:47402", + "right": "issue:46931", + "accept": false, + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they describe the same request." + }, + { + "left": "issue:44370", + "right": "issue:46931", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:44365", + "right": "issue:46931", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:43834", + "right": "issue:46931", "accept": false, - "reason": "Different visible title content and no explicit linkage." + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." }, { "left": "issue:46931", "right": "issue:47402", "accept": false, - "reason": "Different visible content; not the same underlying issue." + "reason": "Different titles and no evidence of the same underlying translation request." }, { "left": "issue:47299", "right": "issue:47402", "accept": false, - "reason": "The titles differ and there is no strong evidence of a single shared issue." + "reason": "Different titles and no evidence of the same underlying translation request." + }, + { + "left": "issue:46226", + "right": "issue:47299", + "accept": false, + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they represent the same request." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The exact-match placeholder translation issues (43834/44365/44370/47402) are reasonably treated as duplicates based on identical titles and very high similarity, while the language-specific translation issues are kept separate. The Flash Attention 2 and GGUF issues are correctly distinguished as different feature requests. No soft-edge verdict appears overstated." - }, - "evaluator_used": true, + "evaluator_result": null, + "evaluator_used": false, "retried": false, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -21679,12 +21767,12 @@ "soft_pairs": [ "issue:30396|issue:31103" ], - "prepared_review_unit_hash": "fadbe6e53619f8fe79dba52e993743f309b3acf3ff0c1b42113b13e2b4723f21", + "prepared_review_unit_hash": "37855664083aa72d66d91c6fdc50d022f0c490c43cfb54cb156b83e23beb38ac", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1540, + "serialized_chars": 1539, "estimated_input_tokens": 385, "estimated_eval_tokens": 1026 }, @@ -21692,7 +21780,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1540, + "serialized_chars": 1539, "estimated_input_tokens": 385, "estimated_eval_tokens": 1026 }, @@ -21701,18 +21789,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are closed feature requests about adding model support, but they target different additions and do not clearly describe the same underlying change. Treat them as related, not duplicates.", - "confidence": 0.79, - "canonical_issue_reason": "Issue 30396 is the more specific and self-contained request; it names a concrete model addition and reads like the clearer original feature request.", + "summary": "These are two highly similar feature requests, but they target different model additions: HelpingAI-3B-v2.2 vs HelpingAIForCausalLM. They are not duplicates of the same underlying change.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #30396 is the clearest standalone representative because its title is more specific about the model being added, but it does not subsume #31103 since they request different artifacts.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 30396 is the best representative for the cluster because it is the more explicit feature request, but it should not be considered a duplicate of 31103.", + "best_issue_reason": "Issue #30396 is the best representative for the cluster only in a loose similarity sense, but the items should not be merged as duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:30396", "right": "issue:31103", "accept": false, - "reason": "Reject: both are model-addition requests in the same broad area, but they name different targets ('HelpingAI-3B-v2.2' vs 'HelpingAIForCausalLM') and there is no evidence they refer to the same concrete change." + "reason": "High textual similarity, but the requested additions are different models and therefore different concrete changes; they should be treated as separate issues." } ] }, @@ -21821,214 +21909,214 @@ "issue:42503|issue:46653", "issue:45027|issue:46653" ], - "prepared_review_unit_hash": "b2fbdd12750b529800e95840a2a614f834a59fa307a987e8208139f3c4447171", + "prepared_review_unit_hash": "0a790c59235f4c232afdd8adb3a5bd21eebf1b77903b59672f29d5cbc47f088c", "component_budget": { "node_count": 23, "item_count": 23, "soft_pair_count": 59, - "serialized_chars": 33608, - "estimated_input_tokens": 8402, - "estimated_eval_tokens": 17060 + "serialized_chars": 33618, + "estimated_input_tokens": 8405, + "estimated_eval_tokens": 17066 }, "budget": { "node_count": 23, "item_count": 23, "soft_pair_count": 59, - "serialized_chars": 33608, - "estimated_input_tokens": 8402, - "estimated_eval_tokens": 17060 + "serialized_chars": 33618, + "estimated_input_tokens": 8405, + "estimated_eval_tokens": 17066 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is a broad mix of unrelated model-request issues. The titles often share generic phrasing like \"Add\" or \"Support,\" but they target different model families and code paths, so none look like true duplicates.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue stands out: the items span unrelated additions for vision, speech, OCR, BERT-like, and multimodal models.", + "summary": "This cluster is mostly a set of distinct model-support/model-addition requests that only look similar because they share template wording. I found no true duplicate issue/PR pairs; the soft links are thematic rather than same-change duplicates.", + "confidence": 0.84, + "canonical_issue_reason": "issue:37334 is the most central and broadly phrased request in the set, with several soft links and discussion activity, so it is the best representative issue for this cluster's theme.", "canonical_pr_reason": null, - "best_issue_reason": "No global best issue for deduping; any one ticket would be too narrow to represent the rest of the cluster.", + "best_issue_reason": "If one issue must stand in for the cluster, issue:37334 is the best fit because it is a broad model-support request and sits near the middle of the similarity graph; still, it should not be treated as a duplicate umbrella for the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36718", "right": "issue:47004", "accept": false, - "reason": "RoMa keypoint matcher vs LoMa keypoint detector/matcher are different model requests, not the same change." + "reason": "Related domain (keypoint matching), but RoMa vs LoMa are different models and different requests." }, { "left": "issue:36718", "right": "issue:41115", "accept": false, - "reason": "Keypoint matcher request and MiniCPM3 model architecture request are unrelated." + "reason": "Different model families and tasks: RoMa keypoint matcher vs MiniCPM3 model architecture." }, { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "EuroBert config support and DEIMv2 model support are different model additions." + "reason": "Both are model-add requests, but EuroBert and DEIMv2 are unrelated model families." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBert config work is unrelated to Qwen3-TTS support." + "reason": "EuroBert config support and Qwen3-TTS support are different model requests." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "EuroBert and RF-DETR are different model families and different code paths." + "reason": "EuroBert config vs RF-DETR model support are unrelated additions." }, { "left": "issue:36817", "right": "issue:42954", "accept": false, - "reason": "EuroBert config and MiMo are unrelated requests." + "reason": "EuroBert and XiaoMi MiMo are different model families with no same underlying change." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are separate detection models, not one concrete bug/change." + "reason": "RF-DETR and DEIMv2 are distinct object-detection model requests." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "RF-DETR model support and LFM2.5 Audio support are unrelated." + "reason": "RF-DETR and LFM2.5 Audio are different modalities and different implementations." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR model addition and weight-sparse transformer support are different changes." + "reason": "RF-DETR model support is unrelated to a weight-sparse transformer request." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR and Qwen3-ASR are different model requests." + "reason": "RF-DETR and Qwen3-ASR target different model families and code paths." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR and Qwen3-TTS are unrelated model additions." + "reason": "RF-DETR vision model support is unrelated to Qwen3-TTS support." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR target different architectures and tasks." + "reason": "NeoBERT and Qwen3-ASR are different model additions, not the same change." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT support and weight-sparse transformer support are not the same concrete change." + "reason": "NeoBERT support and weight-sparse transformer support are different implementation requests." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT and LFM2.5 Audio are unrelated model requests." + "reason": "NeoBERT and LFM2.5 Audio are different model families and tasks." }, { "left": "issue:37015", "right": "issue:37334", "accept": false, - "reason": "NeoBERT and LLMDet are different model families." + "reason": "NeoBERT and LLMDet are distinct model requests; only the request template is similar." }, { "left": "issue:36879", "right": "issue:37015", "accept": false, - "reason": "RF-DETR and NeoBERT are unrelated additions." + "reason": "RF-DETR and NeoBERT are unrelated model additions." }, { "left": "issue:37334", "right": "issue:41115", "accept": false, - "reason": "LLMDet support and MiniCPM3 architecture support are different model changes." + "reason": "LLMDet support and MiniCPM3 architecture support are different models and scopes." }, { "left": "issue:37334", "right": "issue:43837", "accept": false, - "reason": "LLMDet and Qwen3-ASR are unrelated." + "reason": "LLMDet support and Qwen3-ASR support are separate model requests." }, { "left": "issue:37334", "right": "issue:43671", "accept": false, - "reason": "LLMDet and Qwen3-TTS are different model requests." + "reason": "LLMDet and Qwen3-TTS are different model families and code paths." }, { "left": "issue:37334", "right": "issue:42503", "accept": false, - "reason": "LLMDet and ModernVBERT are distinct model additions." + "reason": "LLMDet support and ModernVBERT model support are unrelated additions." }, { "left": "issue:40330", "right": "issue:42503", "accept": false, - "reason": "open-qwen2vl-base and ModernVBERT are different models with different implementation needs." + "reason": "open-qwen2vl-base and ModernVBERT are different models." }, { "left": "issue:40330", "right": "issue:43671", "accept": false, - "reason": "open-qwen2vl-base and Qwen3-TTS are unrelated requests." + "reason": "open-qwen2vl-base and Qwen3-TTS are unrelated model requests." }, { "left": "issue:40330", "right": "issue:42954", "accept": false, - "reason": "open-qwen2vl-base and MiMo are not the same underlying model change." + "reason": "open-qwen2vl-base and XiaoMi MiMo are different model families." }, { "left": "issue:40330", "right": "issue:46653", "accept": false, - "reason": "open-qwen2vl-base is a model request; Mirko Privitera is not the same kind of artifact." + "reason": "open-qwen2vl-base is a model request; Mirko Privitera is unrelated." }, { "left": "issue:40330", "right": "issue:42738", "accept": false, - "reason": "open-qwen2vl-base and BERT-like models with RoPE are unrelated." + "reason": "open-qwen2vl-base and BERT-like models with RoPE are not the same change." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 and Qwen3-ASR are different model families." + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are different model additions." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 support and weight-sparse transformer support are different code paths." + "reason": "MiniCPM3 and OpenAI weight-sparse transformer are unrelated requests." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are separate model additions." + "reason": "MiniCPM3 architecture support and ModernVBERT support are separate model additions." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are unrelated model requests." + "reason": "DEIMv2 and Qwen3-TTS are different model families and tasks." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are different BERT-like models, not the same change." + "reason": "NeoBERT and ModernVBERT are different model requests." }, { "left": "issue:42503", @@ -22036,161 +22124,125 @@ "accept": false, "reason": "ModernVBERT and Qwen3-ASR are unrelated model additions." }, - { - "left": "issue:42738", - "right": "issue:45027", - "accept": false, - "reason": "BERT-like models with RoPE and Voxtral-4B-TTS support are different concrete tasks." - }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models and BERT-like models with RoPE are related only loosely, but not the same issue." - }, - { - "left": "issue:42738", - "right": "issue:42954", - "accept": false, - "reason": "BERT-like models with RoPE and MiMo are different requests." - }, - { - "left": "issue:42738", - "right": "issue:46653", - "accept": false, - "reason": "BERT-like models with RoPE and Mirko Privitera are unrelated." - }, - { - "left": "issue:42954", - "right": "issue:44205", - "accept": false, - "reason": "MiMo and SAM3-LiteText are different model families." - }, - { - "left": "issue:37334", - "right": "issue:42954", - "accept": false, - "reason": "LLMDet and MiMo are unrelated model requests." - }, - { - "left": "issue:42954", - "right": "issue:43671", - "accept": false, - "reason": "MiMo and Qwen3-TTS are different models and tasks." + "reason": "ModernVBERT support and BERT-like models with RoPE are related only at a broad language-model level, not the same underlying change." }, { "left": "issue:42503", "right": "issue:42954", "accept": false, - "reason": "ModernVBERT and MiMo are unrelated additions." + "reason": "ModernVBERT and XiaoMi MiMo are different model families." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are both speech-related, but they are different concrete models and changes." + "reason": "Qwen3-TTS and Qwen3-ASR are different concrete support requests." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and Voxtral-4B-TTS-2603 are different TTS model requests, not one mergeable PR." + "reason": "Qwen3-TTS and Voxtral-4B-TTS-2603 are different TTS models, not the same bug/change." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated." + "reason": "Qwen3-TTS and DeepSeek-OCR2 are different models and different tasks." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS support and weight-sparse transformer support are different code paths." + "reason": "Qwen3-TTS support is unrelated to the weight-sparse transformer request." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are different audio model requests." + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio-model requests." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR and EUPE are unrelated model additions." + "reason": "Qwen3-ASR and EUPE are different model families and purposes." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR support and weight-sparse transformer support are not the same change." + "reason": "Qwen3-ASR support and weight-sparse transformer support are different changes." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio and SAM3-LiteText are different model families." + "reason": "LFM2.5 Audio and SAM3-LiteText are different model-support requests." }, { "left": "issue:43909", "right": "issue:46645", "accept": false, - "reason": "LFM2.5 Audio support and Granite SWA+sink option are unrelated." + "reason": "LFM2.5 Audio support and Granite SWA+sink options are different code paths." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Weight-sparse transformer support and EUPE support are different requests." + "reason": "Weight-sparse transformer support and EUPE are unrelated model requests." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 and Voxtral TTS are unrelated model additions." + "reason": "MiniCPM3 architecture support and Voxtral TTS support are different model additions." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral-4B-TTS are both speech-related, but not the same concrete change." + "reason": "Qwen3-ASR and Voxtral TTS are separate audio-model requests." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral-4B-TTS are unrelated." + "reason": "NeoBERT and Voxtral TTS are different model families and tasks." }, { "left": "issue:37334", "right": "issue:45027", "accept": false, - "reason": "LLMDet and Voxtral-4B-TTS are different model requests." + "reason": "LLMDet and Voxtral TTS are unrelated support requests." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR and DeepSeek-OCR2 are unrelated." + "reason": "Qwen3-ASR and DeepSeek-OCR2 are different tasks and models." }, { "left": "issue:42954", "right": "issue:45177", "accept": false, - "reason": "MiMo and DeepSeek-OCR2 are different model families." + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are unrelated model requests." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are separate model requests, not one underlying bug/change." + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." }, { "left": "issue:42954", "right": "issue:46653", "accept": false, - "reason": "MiMo and Mirko Privitera are not the same artifact or change." + "reason": "XiaoMi MiMo and Mirko Privitera are unrelated." }, { "left": "issue:42503", @@ -22202,7 +22254,7 @@ "left": "issue:45027", "right": "issue:46653", "accept": false, - "reason": "Voxtral-4B-TTS-2603 support and Mirko Privitera are unrelated." + "reason": "Voxtral TTS support and Mirko Privitera are unrelated." } ] }, @@ -22232,7 +22284,7 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "423cd57beada2bd06843beea88ad75138edccd860300ba8ec996f88acd5a0aad", + "prepared_review_unit_hash": "16fd8b7c26c997f1cfc85b438dc5abf6d92534a539ca69a559fa517723ec4f9a", "component_budget": { "node_count": 2, "item_count": 2, @@ -22254,18 +22306,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are unrelated: one is a docs request about adding Pruna as an optimization framework, and the other is a Vision RoPE refactor technical task. No duplicate relationship is indicated.", + "summary": "These two issues are unrelated: one is a docs request for adding `pruna` to optimization framework docs, while the other is a Vision RoPE refactor in the model code path.", "confidence": 0.98, - "canonical_issue_reason": "Issue #46443 is the better representative because it is an active technical issue with a concrete implementation focus, while #38740 is a separate closed docs request.", + "canonical_issue_reason": "Issue 46443 is the better anchor because it is a concrete technical refactor in the Vision RoPE code path; issue 38740 is a separate documentation request and does not represent the same underlying problem space.", "canonical_pr_reason": null, - "best_issue_reason": "#46443 best fits as the cluster representative if one must be chosen, since it is the only open engineering issue and not just documentation.", + "best_issue_reason": "Issue 46443 is the strongest standalone issue in this set because it describes a specific implementation change with a clear technical scope, whereas issue 38740 is docs-only and unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Different topics and scopes: documentation about an optimization framework vs a Vision RoPE refactor. They do not appear to describe the same underlying bug or change." + "reason": "Reject: the titles and scopes do not match. One is a documentation feature request, the other is a Vision RoPE refactor; there are no shared files, targets, or evidence of the same underlying bug/change." } ] }, @@ -22295,12 +22347,12 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "0173c9a9baf6be76d12c213d08757eb12faf1eda9cfa9da9f969b93d24c9700b", + "prepared_review_unit_hash": "44c4da4da1a1a051af77626529c66e2d963165de3c347158541571623d878332", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -22308,7 +22360,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1479, + "serialized_chars": 1480, "estimated_input_tokens": 370, "estimated_eval_tokens": 996 }, @@ -22317,18 +22369,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues appear unrelated: one is about block decoding, the other is a request for DeepSeek v3.2 support. I would not cluster them as duplicates.", - "confidence": 0.97, + "summary": "These two issues are not duplicates: one is about block decoding, the other about DeepSeek v3.2 model support.", + "confidence": 0.96, "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 43037 is the more actionable representative because it names a specific model-version support request; issue 41084 is a separate decoding-related topic.", + "best_issue_reason": "Neither issue is a good canonical for the other because they describe different requests; if forced to pick the more concrete standalone item, 41084 is the narrower technical topic.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different underlying problems: block decoding vs. model support for DeepSeek v3.2. They do not describe the same bug or change." + "reason": "Different subjects with no shared target or code path: block decoding vs DeepSeek v3.2 support." } ] }, @@ -22358,40 +22410,40 @@ "soft_pairs": [ "issue:41306|issue:41826" ], - "prepared_review_unit_hash": "c7da709544572cb879440c6f58f8a04b812e699f8edab19093cc1f77b531f78e", + "prepared_review_unit_hash": "05b05c51f6bbdabb64fef6c81d9c4c202d229cadc54c93eb7aebf76a20cf15f5", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1585, - "estimated_input_tokens": 397, - "estimated_eval_tokens": 1050 + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1585, - "estimated_input_tokens": 397, - "estimated_eval_tokens": 1050 + "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 unrelated: one is about enabling a `from_pretrained()` flag for TiledFusedLogitsLoss/FLCE, while the other is a feature request to integrate TiledMLP for reduced memory use. No duplicate or same-underlying-change signal is present.", - "confidence": 0.98, - "canonical_issue_reason": "issue:41306 is the more concrete, narrowly scoped bug/feature request and is easier to treat as the representative issue for this cluster.", + "summary": "The two issues are about different features in different model components: one requests a `from_pretrained()` flag for TiledFusedLogitsLoss/LigerFusedLinearCrossEntropy, while the other asks for integrating TiledMLP to reduce memory footprint. They are not duplicate reports of the same bug or change.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:41306 is the strongest standalone issue because it targets a specific API behavior change, whereas issue:41826 is broader and more speculative.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41306", "right": "issue:41826", "accept": false, - "reason": "Different features and different code paths; they do not describe the same underlying bug or change." + "reason": "Different underlying requests: one is about a pretrained-loading flag for FLCE/TiledFusedLogitsLoss, the other is about adding TiledMLP for memory reduction. They touch different functionality and would not be merged as one issue." } ] }, @@ -22421,12 +22473,12 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "42e14c8cae58555f6fc1f8c5a643e7a4a588ba5c7f32664cbc24be1886befbd0", + "prepared_review_unit_hash": "34ddaaacba978db0395b335ab6fc33076ba0452d358da06a8d1288ef2876faca", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -22434,7 +22486,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1583, + "serialized_chars": 1584, "estimated_input_tokens": 396, "estimated_eval_tokens": 1048 }, @@ -22443,18 +22495,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues are related only at a broad Trainer/customization level, but they describe different problems and code paths: one is about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers.", - "confidence": 0.91, - "canonical_issue_reason": "Issue #43598 is the more concrete Trainer behavior change request, centered on `remove_unused_column` and customizability.", + "summary": "These are two unrelated Trainer customization issues: one about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "Neither issue is clearly canonical for the other; they address different Trainer behaviors and different code paths.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43598 is the best representative because it names a specific Trainer mechanism and likely corresponds to a distinct actionable fix.", + "best_issue_reason": "Issue #43598 is the more specific customization concern, but it is not a duplicate of #43600 and should not be treated as a representative for that topic.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "They are not the same underlying bug: one concerns unused-column removal in Trainer, while the other concerns FLOPS estimation for custom trainers. Shared Trainer/customization context is too broad to merge." + "reason": "Different underlying problems: one concerns unused-column removal/customizability in Trainer, the other concerns FLOPS estimation for custom trainers. Same broad area, but not the same bug or change." } ] }, @@ -22484,12 +22536,12 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "1015f5e46f51136e1ded88ac9016c8b95f7375742a4c9f755868d964b510985a", + "prepared_review_unit_hash": "468f7d5c3d1cea4e7dd16b9c4ae7a6b045b6f7b94c8dad045b3d73bae2b3e522", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1609, + "serialized_chars": 1611, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -22497,7 +22549,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1609, + "serialized_chars": 1611, "estimated_input_tokens": 403, "estimated_eval_tokens": 1062 }, @@ -22506,18 +22558,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two different feature requests for different Qwen variants and different auto-classification paths, so they should not be merged as duplicates.", - "confidence": 0.93, - "canonical_issue_reason": null, + "summary": "These are two different model-support requests in the Transformers registration layer: one for Qwen3-Omni AutoModel/AutoModelForConditionalGeneration, the other for Qwen3.5 AutoModelForSequenceClassification. They do not describe the same underlying bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43684 is the broader and more general registration request for Qwen3-Omni, so it is the better representative issue if one must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43684 is the broader, more general registration request, but it does not cover 44405\u2019s specific AutoModelForSequenceClassification support for Qwen3.5.", + "best_issue_reason": "Issue 43684 is the stronger global representative because it targets core model registration rather than a narrower downstream sequence-classification variant.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different model families and different capabilities: Qwen3-Omni auto-registration vs Qwen3.5 sequence-classification support. They do not look like the same underlying bug or change." + "reason": "Different model families and different APIs/tasks: Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration vs Qwen3.5 sequence-classification support. Not the same fix." } ] }, @@ -22554,12 +22606,12 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "59ac9588b13b05c1c1bf8a1e01a058abcd90168b9173476cc02f0fec5466dc02", + "prepared_review_unit_hash": "9aeb06d848ea67e4010ef9747519442f1c4eacb7309b65f372d4281deb638c65", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4263, + "serialized_chars": 4264, "estimated_input_tokens": 1066, "estimated_eval_tokens": 2388 }, @@ -22567,7 +22619,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4263, + "serialized_chars": 4264, "estimated_input_tokens": 1066, "estimated_eval_tokens": 2388 }, @@ -22576,54 +22628,54 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Three issues share the same sparse \"huggingface\" title and look like duplicates; #43835 (\u201cFett\u201d) appears unrelated. I would cluster the huggingface issues together and keep the earliest one as canonical.", - "confidence": 0.87, - "canonical_issue_reason": "#43845 is the earliest of the near-identical \"huggingface\" issues and is a reasonable representative for the duplicate cluster.", + "summary": "Three issues are near-duplicates with the same truncated title \"huggingface \"; the \"Fett\" issue looks unrelated. I would collapse the three huggingface tickets together and keep the earliest one as canonical.", + "confidence": 0.92, + "canonical_issue_reason": "issue #43845 is the earliest of the near-identical \"huggingface \" issues, so it is the best canonical representative.", "canonical_pr_reason": null, - "best_issue_reason": "#43845 is the strongest single issue to keep: it is the earliest duplicate-looking \"huggingface\" report and best represents the repeated topic.", + "best_issue_reason": "#43845 is the cleanest cluster anchor: same title as the other two matching issues, earliest creation time, and no conflicting signals.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles; \"Fett\" does not look like the same underlying report as the repeated \"huggingface\" issues." + "reason": "Different titles and no evidence they describe the same underlying problem; \"Fett\" does not match the huggingface cluster." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles; no evidence these describe the same bug or request." + "reason": "Different titles and no shared details beyond weak similarity; insufficient evidence of the same issue." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles; no indication of a shared underlying issue." + "reason": "\"Fett\" appears unrelated to the huggingface tickets and lacks any concrete overlap." }, { "left": "issue:43845", "right": "issue:43846", "accept": true, - "reason": "Same title, same body length, and created one minute apart; these look like duplicates of the same report." + "reason": "Same truncated title \"huggingface \" and no distinguishing body details; these look like duplicate reports of the same issue." }, { "left": "issue:43845", "right": "issue:43859", "accept": true, - "reason": "Same title and clearly the same repeated topic; likely another duplicate in the huggingface cluster." + "reason": "Same title and no differing context provided, so they plausibly refer to the same underlying issue." }, { "left": "issue:43846", "right": "issue:43859", "accept": true, - "reason": "Same title and repeated topic; despite the shorter body, it still looks like the same underlying issue." + "reason": "Same title and no differentiating information; these appear to be duplicates." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary stays within the packet: #43835 has a different title and the three 'huggingface' issues share the same title with soft-similarity evidence, so clustering them together and treating #43845 as the canonical representative is reasonable. The soft-edge accepts are also conservative enough given the identical titles and high similarity scores." + "feedback": "The summary is grounded in the packet: #43845/#43846/#43859 share the exact truncated title and #43845 is the earliest, while #43835 has a distinct title and much weaker similarity. The duplicate/cluster judgments are conservative enough, and nothing appears overstated." }, "evaluator_used": true, "retried": false, @@ -22665,106 +22717,106 @@ "issue:44573|issue:47036", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "fb1ef3f0fc39eaacd3de782a80ca55bba8c61fd6d122408ee91fcf8bc0747236", + "prepared_review_unit_hash": "cd83e2cbd26e2f5d1b3c61e707b8d206c978c2d863b8c858a2e06168afe1d96d", "component_budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7361, - "estimated_input_tokens": 1841, - "estimated_eval_tokens": 3938 + "serialized_chars": 7352, + "estimated_input_tokens": 1838, + "estimated_eval_tokens": 3932 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 12, - "serialized_chars": 7361, - "estimated_input_tokens": 1841, - "estimated_eval_tokens": 3938 + "serialized_chars": 7352, + "estimated_input_tokens": 1838, + "estimated_eval_tokens": 3932 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are low-information, mostly spam-like issues with no concrete shared bug or change, so none of the soft pairs should be merged as duplicates.", - "confidence": 0.91, - "canonical_issue_reason": null, + "summary": "The cluster is mostly low-information spam or generic placeholder issues with no convincing shared bug/change. The only exact textual duplicate is the pair 44572/44573 (''), but that is still too content-free to confidently treat as the same underlying issue as the rest.", + "confidence": 0.92, + "canonical_issue_reason": "issue:44572 is the earliest of the only exact duplicate pair ('' / '') and is the cleanest representative among otherwise unrelated, low-signal issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44418 is the least spammy and most descriptive title, but it is still too vague to be a reliable canonical representative.", + "best_issue_reason": "issue:44572 is the best available representative because it matches 44573 exactly by title and is the only plausible duplicate anchor in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are just '' with no body detail; no evidence of the same underlying issue beyond generic similarity." + "reason": "Both are placeholder spam, but there is no evidence they describe the same underlying report." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Both are just '' with no body detail; generic title match is not enough to treat them as duplicates." + "reason": "Both are placeholder spam, but the shared label is too generic to prove a duplicate." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is spam, the other is a vague documentation suggestion; no concrete shared bug or change is described." + "reason": "Spam placeholder vs. generic documentation suggestion; unrelated." }, { "left": "issue:44371", "right": "issue:47036", "accept": false, - "reason": "Spam vs. a community benchmark tip are unrelated on their face; no duplicate signal beyond weak text similarity." + "reason": "Spam placeholder vs. benchmark tip; unrelated." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Spam and 'Check out \"Google LLC\"' do not indicate the same problem or request." + "reason": "Spam placeholder vs. 'Check out \"Google LLC\"'; no shared bug/change." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "A vague documentation enhancement and a spam issue are not the same actionable report." + "reason": "Generic documentation suggestion vs. spam placeholder; not the same issue." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "A vague documentation enhancement and a spam issue are not the same actionable report." + "reason": "Generic documentation suggestion vs. spam placeholder; not the same issue." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are '', but there is no substantive description to confirm they report the same underlying item." + "reason": "Exact same placeholder title, but there is still no substantive evidence of a shared underlying bug/change." }, { "left": "issue:44572", "right": "issue:47036", "accept": false, - "reason": "Spam vs. benchmark tip are unrelated; the titles do not identify a shared bug or change." + "reason": "Spam placeholder vs. benchmark tip; unrelated." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Spam and 'Check out \"Google LLC\"' appear to be different low-signal submissions, not duplicates." + "reason": "Spam placeholder vs. unrelated promotional-looking title; no duplicate evidence." }, { "left": "issue:44573", "right": "issue:47036", "accept": false, - "reason": "Spam vs. benchmark tip are unrelated; no concrete overlap in problem statement." + "reason": "Spam placeholder vs. benchmark tip; unrelated." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "Spam and 'Check out \"Google LLC\"' do not describe the same underlying issue." + "reason": "Spam placeholder vs. unrelated promotional-looking title; no duplicate evidence." } ] }, @@ -22794,12 +22846,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "542f7caf103dfcca01676b21efbd9a4c95a85d40d58d5e019b679c1a56d07d80", + "prepared_review_unit_hash": "bb05c247776f48a207eb4a3f108c796c7b3a30d6c9e33ed66a5d83e1bf3ca51a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1646, + "serialized_chars": 1645, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -22807,7 +22859,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1646, + "serialized_chars": 1645, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -22816,28 +22868,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues look strongly similar, but the packet only proves identical titles, close body sizes, and a soft-similarity signal. That is enough for a tentative duplicate suspicion, not a confident same-issue determination.", - "confidence": 0.67, - "canonical_issue_reason": "If this cluster is treated as duplicate issues, #45145 is the better canonical candidate because it has the same title and at least one inbound reference, making it slightly more established.", + "summary": "#45145 and #45151 look like a duplicate issue pair: they have the same title, similar-sized bodies, and a soft-similarity signal. #45145 is the better canonical representative because it has an inbound reference, giving it slightly stronger linkage.", + "confidence": 0.87, + "canonical_issue_reason": "#45145 is the better canonical issue because it has the same title as #45151 and has an inbound reference, so it is the more connected record.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best issue representative for the cluster only in the provisional sense: it is slightly more connected than #45151, but the duplicate relationship itself is not firmly established from the available evidence.", + "best_issue_reason": "#45145 is the best issue to keep as the representative duplicate target: it is tied to the same titled report and has the stronger reference signal.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", - "accept": false, - "reason": "This is a tentative similarity match, not a confirmed duplicate: the evidence is limited to identical titles, similar body length, and soft_similarity, with deterministic_accept=false and no substantive body comparison shown." + "accept": true, + "reason": "Same title plus the provided similarity signal and comparable body sizes make this a plausible duplicate issue pair, so it is reasonable to treat them as the same request." } ] }, "evaluator_result": { "accept": false, - "feedback": "The summary is a bit too strong for the evidence. The packet only shows identical titles, similar body lengths, and a soft-similarity signal; it does not show that the issue bodies are substantively the same. The accepted soft edge is also not conservative here because the pairing is based only on soft_similarity and deterministic_accept is false. Please downgrade the verdict to tentative rather than a clear duplicate." + "feedback": "The pair is plausibly related, but the rationale overstates the evidence. The packet only shows identical titles, similar body lengths, and a soft-similarity signal; it does not provide the issue bodies, so claims of \u201cnear-identical content\u201d or the same underlying request are not directly grounded. The canonical-choice note about #45145 having an inbound reference is supported, but the duplicate-verdict language should be more conservative." }, "evaluator_used": true, "retried": true, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -22914,12 +22966,12 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "117938920d4a664af88445be661a8b740138cd03c2608e1bee30280b7953ffbe", + "prepared_review_unit_hash": "6be3dd2267072abaa1b6435fdb6bb364479f0ba0a4f52e83fddec91fca0684ab", "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 }, @@ -22927,7 +22979,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1475, + "serialized_chars": 1476, "estimated_input_tokens": 369, "estimated_eval_tokens": 994 }, @@ -22936,7 +22988,7 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two closed issues do not look like duplicates or the same underlying bug/report: they have different personal-name titles, different dates, and no bodies or cross-references tying them together.", + "summary": "Two closed issues with person-name-and-date titles and no bodies or links; they do not appear to describe the same bug or request.", "confidence": 0.95, "canonical_issue_reason": null, "canonical_pr_reason": null, @@ -22947,7 +22999,7 @@ "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Titles reference different people and different dates; with no descriptions, references, or shared context, there is no evidence they describe the same issue." + "reason": "Different names and birth dates, with no shared context or evidence of the same underlying issue." } ] }, @@ -22977,7 +23029,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "19e452a54d4b9b979a032cce6d779024f632db1c7d36f516e3e01b26f4e6ed25", + "prepared_review_unit_hash": "f2dc1defd9ac31a3aaf426c826f1fbd2bdc3039f9fc4f59ffb77502a978e17af", "component_budget": { "node_count": 2, "item_count": 2, @@ -22999,18 +23051,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues appear unrelated: they are just personal-name/date titles with no body, targets, or shared evidence of a common bug report. No duplicate relationship is supported.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "The two issues do not describe the same bug or request; they look like unrelated name/date entries rather than duplicate technical reports.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the titles are distinct and contain no shared problem description, repro, or target area.", "canonical_pr_reason": null, - "best_issue_reason": "Neither issue is a good canonical choice because both lack any substantive report content, and they do not describe the same problem.", + "best_issue_reason": "Neither issue is a good canonical representative because neither contains substantive issue content to anchor deduplication.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "Only soft similarity is present; the titles refer to different names and dates, with no shared issue content or explicit linkage indicating the same underlying bug." + "reason": "Reject: the titles are different personal-name/date strings with no evidence of the same underlying bug or change." } ] }, @@ -23040,7 +23092,7 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "a19c4ff53b4ad9fec42764f6251da5b15d13a2fcb2afe08af95cb66c9db14c87", + "prepared_review_unit_hash": "54d0c733f1cfdab185583f480a52470e4db181e20a635c16121b782240cdd940", "component_budget": { "node_count": 2, "item_count": 2, @@ -23062,18 +23114,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are not duplicates: one requests a new Qwen3_5 token-classification/value-model class, while the other asks for FP8 training support. They concern different features and code paths.", - "confidence": 0.96, - "canonical_issue_reason": "issue:45810 is the more concrete, narrowly scoped feature request; it is a better representative if one issue must be kept as the cluster anchor.", + "summary": "These are unrelated feature requests: one asks for Qwen3_5 token classification/value-model support, the other asks for FP8 training support. They do not describe the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45810 is the best issue overall because it is specific and actionable, whereas issue:46295 is a broad training-capability request.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different underlying requests: adding a Qwen3_5ForTokenClassification/value-model class versus enabling FP8 training support. They do not describe the same bug or change." + "reason": "Different feature areas and different requested changes; no shared code-path or concrete duplicate signal." } ] }, @@ -23113,91 +23165,91 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "838af35a8cac4a3f4cb7ef7a00f5a46821847b1b38015011eafcb615bcc2891e", + "prepared_review_unit_hash": "6316f67bca0c8a7564076b12585665b9a35f4f67468604d4d29267440b50d985", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5725, - "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": 5725, - "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 set is not one single duplicate cluster. It contains two exact duplicate issue pairs (45870/45871 and 45872/45873) plus an unrelated dataset issue (45869). 45870 is the cleanest representative of the exact duplicate \"Example: Financial time series prediction with DataForge\" pair, but the broader set should be split into separate groups.", - "confidence": 0.95, - "canonical_issue_reason": "45870 is an exact duplicate of 45871 and is the earliest, clearest representative of that duplicate pair.", + "summary": "This cluster appears to contain two separate likely-duplicate issue pairs: 45870/45871 and 45872/45873. The evidence shown is strong but not definitive, based mainly on identical titles and matching body lengths. 45869 looks related by domain, but it is a separate dataset promo/request rather than the same underlying issue.", + "confidence": 0.82, + "canonical_issue_reason": "No single canonical issue cleanly covers the whole cluster, since it likely splits into two separate near-duplicate subgroups. If a representative must be chosen for the first subgroup, 45870 is the best anchor because it is the earlier of the near-identical 45870/45871 pair; 45871 looks like a likely duplicate.", "canonical_pr_reason": null, - "best_issue_reason": "45870 is the best single issue to keep as the representative for the exact duplicate example-prediction pair.", + "best_issue_reason": "45870 is the best overall issue representative here: it is early, self-contained, and part of a strong duplicate-candidate pair. 45869 is a different promotional dataset request, and 45872/45873 form a separate likely-duplicate subgroup.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Different topics: a financial dataset bundle versus an example about time-series prediction." + "reason": "Related financial/DataForge context, but a dataset promo/request is not the same underlying issue as an example prediction task." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Different topics: a financial dataset bundle versus an example about time-series prediction." + "reason": "Same mismatch as 45869/45870: shared domain, but different request type and scope." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Same title and apparently the same example issue; this looks like an exact duplicate." + "reason": "Identical title and matching body length make this a strong duplicate candidate." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Related broad domain, but different concrete issue: prediction example versus scraping-service topic." + "reason": "Both mention DataForge/financial time series, but one is an example request and the other is a scraping-service request; different underlying ask." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Related broad domain, but different concrete issue: prediction example versus scraping-service topic." + "reason": "Same as 45870/45872: related theme, but not the same issue." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Different concrete issue themes; these are not the same underlying bug or request." + "reason": "Different concrete request even though both are in the same general DataForge area." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Different concrete issue themes; these are not the same underlying bug or request." + "reason": "Related context only; the issue themes do not match closely enough to be duplicates." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Same title and likely the same request; this looks like an exact duplicate." + "reason": "Identical title and matching body length make this a strong duplicate candidate." } ] }, "evaluator_result": { - "accept": true, - "feedback": "Grounded and conservative overall: 45870/45871 and 45872/45873 have identical titles and matching body lengths, while 45869 is clearly a different topic. The split into two duplicate pairs plus one unrelated issue is supported by the packet." + "accept": false, + "feedback": "The clustering is plausible, but the writeup overstates the evidence. The packet only shows identical titles and matching body lengths for 45870/45871 and 45872/45873, which supports a likely-duplicate reading, not a definitive 'exact duplicate' claim. The summary and canonical reasons should be softened to conservative language (e.g. 'likely duplicate' or 'strong duplicate candidate'). 45869 being separate from the DataForge example pair is grounded, but the duplicate certainty is too strong for the available evidence." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -23221,7 +23273,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "809f1a65718778fb896c0dff500e16f665217fce66df67fe27cbc9134c6d4275", + "prepared_review_unit_hash": "074c14a904ec0d17a19b58d32b376d540c6f975ac4d342f8e83f95d5dccee838", "component_budget": { "node_count": 2, "item_count": 2, @@ -23243,27 +23295,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Issue 46439 remains the best canonical issue because it describes the broader root cause, while 46441 reads as a narrower symptom report. The pair is plausibly related through `cache_implementation=\"static\"` and linear-attention behavior, but the evidence is too weak to confirm they are the same underlying bug.", - "confidence": 0.77, - "canonical_issue_reason": "46439 is the better canonical issue because it names the likely underlying mechanism (`early_initialization` corrupting linear-attention layers) instead of just one downstream failure mode.", + "summary": "These two issues are related to static-cache failures in linear-attention/Qwen3.5 paths, but they describe different failure modes and are not clearly the same underlying bug, so I would not deduplicate them.", + "confidence": 0.74, + "canonical_issue_reason": "Issue #46439 is the better canonical candidate because it states a broader, more root-cause-oriented problem: `early_initialization` corrupts linear-attention layers and leads to shape/runtime failures.", "canonical_pr_reason": null, - "best_issue_reason": "46439 is the strongest cluster representative since it is the most root-cause-oriented report and better captures the shared failure family.", + "best_issue_reason": "#46439 best captures the general static-cache/linear-attention corruption issue; #46441 looks like a narrower downstream symptom (`KeyError: 'linear_attention'`) on one model.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "The reports are similar, but the packet only provides `soft_similarity` with no shared targets/files and `deterministic_accept: false`. That supports a possible duplicate relationship, not a confirmed same-bug match." + "reason": "Both mention static cache and linear attention, but one is a shape/runtime corruption bug and the other is a `KeyError` in Qwen3.5. That is too weak to treat as the same concrete underlying bug." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The canonical choice of 46439 is reasonable, but the soft-edge accept is too strong for the evidence in the packet. The only pair evidence is `soft_similarity` with no shared targets/files and `deterministic_accept: false`, so saying both issues \"point to the same\" failure path overstates the support. This should be phrased more cautiously as a possible or likely match, not a confirmed one." - }, - "evaluator_used": true, - "retried": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -23287,7 +23336,7 @@ "soft_pairs": [ "issue:46551|issue:46552" ], - "prepared_review_unit_hash": "58d22273b29feb5ba59d4f56d04083a157c31742df08fd4cdad4b3cf3edb5d81", + "prepared_review_unit_hash": "83abc53c47505a1a7dd502b3f09f99e3713afad9dcaf7583ea842f32e7bd06b0", "component_budget": { "node_count": 2, "item_count": 2, @@ -23309,24 +23358,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are duplicate issue reports with identical titles and bodies about Wav2Vec2CTCTokenizer v5 misclassifying the word delimiter as a special token. Issue #46552 is the better canonical because it already has inbound references and likely serves as the main thread.", + "summary": "The two issues are exact duplicates: identical titles and apparently identical bodies, describing the same Wav2Vec2CTCTokenizer v5 special-token regression.", "confidence": 0.99, - "canonical_issue_reason": "#46552 has the same report as #46551 but with more linkage/inbound references, making it the better central issue to keep.", + "canonical_issue_reason": "issue #46552 is the better canonical issue because it already has inbound references, suggesting it is the more visible/linked duplicate target.", "canonical_pr_reason": null, - "best_issue_reason": "#46552 is the best issue for the cluster because it is the same bug report and appears to be the more established discussion point.", + "best_issue_reason": "#46552 is the clearest single issue to keep because it fully matches the bug description and appears to be the more referenced record.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46551", "right": "issue:46552", "accept": true, - "reason": "Identical title and body describe the same tokenizer bug, so this is a clear duplicate issue pair." + "reason": "Exact duplicate match: same title and same underlying tokenizer bug report, with no distinguishing details between them." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusion is grounded in the packet: both issues have the same title, the same body length, and maximum soft-similarity, so treating them as duplicates is reasonable. The canonical choice of #46552 is also supported by its inbound references. The only slightly speculative phrase is \u2018likely serves as the main thread,\u2019 but it does not materially overstate the evidence." + "feedback": "The decision is well grounded: both issues have the same title, the same body length, and a perfect similarity signal, while #46552 also has inbound references. The duplicate/canonical choice is conservative and consistent with the packet." }, "evaluator_used": true, "retried": false, @@ -23353,7 +23402,7 @@ "soft_pairs": [ "issue:46657|issue:46692" ], - "prepared_review_unit_hash": "70871a11ab701cbda34cb2b92bbf0656ad9540aa195d54ab1ef73e384d0a8c21", + "prepared_review_unit_hash": "3f744fdbfdec1f4f8ff4180080ec6d035ea41d61fd41dcc757de037ed8b2f08e", "component_budget": { "node_count": 2, "item_count": 2, @@ -23375,18 +23424,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues are both about training object detection models, but they report different failure modes: a tensor size mismatch vs. an out-of-memory crash after several epochs. They do not look like the same underlying bug, so they should not be deduplicated.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "The two issues are both about training object detection models, but they describe different failure modes: a tensor size mismatch versus out-of-memory after several epochs. They do not look like the same underlying bug.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #46657 is the more concrete, narrowly defined bug report with a specific tensor-shape mismatch during training, making it a better representative of the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #46657 is the more concrete, specific bug report, but it still does not represent the other issue\u2019s OOM behavior.", + "best_issue_reason": "Issue #46657 is the best standalone issue because it describes a specific reproducible error path, unlike #46692 which is a broader resource-exhaustion symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46657", "right": "issue:46692", "accept": false, - "reason": "The symptoms are materially different: one is a tensor-shape mismatch during training, the other is memory exhaustion after some epochs. Shared domain alone is not enough to treat them as the same bug." + "reason": "Different underlying problems: one is a tensor dimension mismatch, the other is training-time OOM after some epochs. Similar domain, but not the same bug or change." } ] }, @@ -23416,7 +23465,7 @@ "soft_pairs": [ "issue:46822|issue:46823" ], - "prepared_review_unit_hash": "7a41c7d4bc0b50ceaf741389cfd72d2ab2fc899078be7d366ba11adc544f3832", + "prepared_review_unit_hash": "0cb45966323771665dca03a7ad04a6de119c01f0a1e4338935f76998bd219855", "component_budget": { "node_count": 2, "item_count": 2, @@ -23438,18 +23487,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two different model-loading/memory bugs: one about ZeRO-3 zero.Init failing to partition a composite submodule and causing OOM, the other about 4-bit auto device mapping incorrectly spilling to CPU and raising a ValueError. They are not duplicates and should stay separate.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "These two issues describe different loading bugs in different code paths: one is ZeRO-3 zero.Init failing to partition a composite submodule and causing OOM, the other is 4-bit + device_map/max_memory incorrectly dispatching a model to CPU and raising ValueError. They should not be deduplicated.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 46822 is the better representative issue for the cluster because it has a clearer, single root-cause description tied to a concrete partitioning failure during multi-GPU load.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 46822 is the stronger standalone report: it names the exact affected mechanism (ZeRO-3 zero.Init), the failing submodule, and the observed symptom (OOM), making it more actionable.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46822", "right": "issue:46823", "accept": false, - "reason": "Different failure modes and code paths: ZeRO-3 submodule partitioning/OOM vs 4-bit device_map/max_memory CPU dispatch error; only a broad loading-memory similarity." + "reason": "Different underlying bugs and code paths: ZeRO-3 zero.Init partitioning/OOM versus 4-bit device_map/max_memory CPU dispatch/ValueError. No shared target or plausible single fix." } ] }, @@ -23553,8 +23602,8 @@ "issue:47227|issue:47377", "issue:47227|issue:47256", "issue:47182|issue:47309", - "issue:47309|issue:47419", "issue:47227|issue:47279", + "issue:47309|issue:47419", "issue:47182|issue:47227", "issue:47227|issue:47265", "issue:47419|issue:47442", @@ -23571,7 +23620,7 @@ "issue:47377|issue:47406", "issue:47279|issue:47406" ], - "prepared_review_unit_hash": "9f896b88b968364f2aeb415a0d368cef41810a1ba6915e8d7b81c948b9ea0103", + "prepared_review_unit_hash": "ef2f6af243a632cbfb02d77cd339768de571326a2597740ddd6c932ee7fbbdf8", "component_budget": { "node_count": 23, "item_count": 23, @@ -23593,438 +23642,438 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are a chronological series of dated '[serge] integration failure triage' issues. They look like separate daily triage threads with the same template, not duplicate reports of one underlying bug. The open 2026-07-20 issue is the best representative for the cluster.", - "confidence": 0.97, - "canonical_issue_reason": "If a single representative is needed, issue:47442 is the best choice because it is the latest open triage thread and has the most recent activity. However, the set appears to be date-specific triage issues rather than true duplicates.", + "summary": "This cluster is a recurring series of dated \"[serge] integration failure triage\" issues, not clear duplicates of one bug. The best anchor is the latest open thread, #47442, which appears to be the active continuation of the series. No PRs are present.", + "confidence": 0.99, + "canonical_issue_reason": "#47442 is the latest open triage thread and has the most current inbound activity, so it is the best anchor for this ongoing series.", "canonical_pr_reason": null, - "best_issue_reason": "Issue:47442 is the best global representative: it is open, newest, and appears to continue the same triage stream as the earlier dated issues.", + "best_issue_reason": "#47442 is the strongest representative of the cluster because it is open, current, and appears to continue the same triage workflow.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46906", "right": "issue:46940", "accept": false, - "reason": "Same triage template, but different dated issue threads; not the same underlying report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46906", "right": "issue:46948", "accept": false, - "reason": "These are separate daily triage issues, not duplicate bug reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46906", "right": "issue:46969", "accept": false, - "reason": "Different dates and likely different daily failure snapshots; do not merge." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46906", "right": "issue:46985", "accept": false, - "reason": "Same series, but each issue is date-specific triage rather than one duplicate." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46906", "right": "issue:47037", "accept": false, - "reason": "The titles match, but the dated triage threads are separate artifacts." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46940", "right": "issue:46948", "accept": false, - "reason": "Adjacent daily triage issues; same format, but not the same underlying issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46948", "right": "issue:46969", "accept": false, - "reason": "Different daily triage threads, not duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46948", "right": "issue:46985", "accept": false, - "reason": "Same triage series, but the date-stamped reports are distinct." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46940", "right": "issue:46969", "accept": false, - "reason": "Separate dated triage issues; no evidence they are the same bug." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46940", "right": "issue:46985", "accept": false, - "reason": "Daily triage threads for different days; reject as duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47037", "right": "issue:47049", "accept": false, - "reason": "Same subsystem and title pattern, but different day-specific triage reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46948", "right": "issue:46998", "accept": false, - "reason": "These look like separate daily integration-failure triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47049", "right": "issue:47053", "accept": false, - "reason": "Distinct dated triage issues, not one duplicated report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46998", "right": "issue:47049", "accept": false, - "reason": "Different daily triage entries; no shared underlying bug shown." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46969", "right": "issue:46985", "accept": false, - "reason": "Same recurring triage pattern, but each issue corresponds to a different date." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46940", "right": "issue:46998", "accept": false, - "reason": "Date-specific triage threads; the similarity is template-level only." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46985", "right": "issue:47049", "accept": false, - "reason": "Separate daily reports, not a single duplicate issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47049", "right": "issue:47091", "accept": false, - "reason": "Same triage series, but different day-specific issue instances." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46998", "right": "issue:47037", "accept": false, - "reason": "Different dated triage threads; not the same concrete problem instance." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46985", "right": "issue:47037", "accept": false, - "reason": "No evidence of one underlying duplicated report beyond the shared template." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46969", "right": "issue:47037", "accept": false, - "reason": "Separate daily triage issues; reject as duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47037", "right": "issue:47091", "accept": false, - "reason": "Different date-stamped triage threads." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47037", + "right": "issue:47053", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46969", "right": "issue:47091", "accept": false, - "reason": "Same series, but the issues represent different days." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47053", "right": "issue:47091", "accept": false, - "reason": "These are distinct triage reports with the same naming pattern." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47091", "right": "issue:47102", "accept": false, - "reason": "Adjacent daily triage entries, but not the same underlying bug." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46998", "right": "issue:47053", "accept": false, - "reason": "Different day-specific triage issues; no duplicate signal." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47053", "right": "issue:47102", "accept": false, - "reason": "Same recurring triage thread format, but separate issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46985", "right": "issue:46998", "accept": false, - "reason": "Different dated triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46998", "right": "issue:47091", "accept": false, - "reason": "Template similarity only; these are separate daily reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47037", "right": "issue:47102", "accept": false, - "reason": "Distinct daily triage issues, not one mergeable duplicate." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46985", "right": "issue:47091", "accept": false, - "reason": "Different dates and likely different failure snapshots." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:46985", "right": "issue:47102", "accept": false, - "reason": "Same series, but each issue is a separate daily triage thread." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47049", "right": "issue:47102", "accept": false, - "reason": "No concrete evidence these are the same bug report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47265", "right": "issue:47279", "accept": false, - "reason": "These are consecutive daily triage issues, not duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47265", "right": "issue:47309", "accept": false, - "reason": "Same named triage stream, but different date-specific issue instances." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47279", "right": "issue:47309", "accept": false, - "reason": "Separate daily triage threads with the same template." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47265", "right": "issue:47377", "accept": false, - "reason": "Different dated triage issues; same series but not the same report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47279", "right": "issue:47377", "accept": false, - "reason": "Each issue is a distinct day\u2019s triage thread." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47256", "right": "issue:47265", "accept": false, - "reason": "Adjacent daily triage issues, not duplicate bug reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47256", "right": "issue:47279", "accept": false, - "reason": "Same recurring triage pattern, but separate artifacts." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47256", "accept": false, - "reason": "These are day-specific triage issues, not the same underlying issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47120", "right": "issue:47256", "accept": false, - "reason": "Different dated integration triage threads; reject as duplicates." - }, - { - "left": "issue:47309", - "right": "issue:47377", - "accept": false, - "reason": "Separate daily triage issues with the same naming scheme." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47102", "right": "issue:47256", "accept": false, - "reason": "Different days in the same triage series, not a single duplicated report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47265", "accept": false, - "reason": "These are distinct daily triage threads." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47120", "right": "issue:47182", "accept": false, - "reason": "Adjacent daily issues, but not the same concrete bug instance." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47120", "right": "issue:47279", "accept": false, - "reason": "Same triage template, different date-stamped issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47120", "right": "issue:47265", "accept": false, - "reason": "Separate daily triage threads rather than duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47120", "right": "issue:47377", "accept": false, - "reason": "Different day-specific integration failure triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47377", "accept": false, - "reason": "Same series, but the issues correspond to different dates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47279", "accept": false, - "reason": "Distinct daily triage reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47227", "right": "issue:47377", "accept": false, - "reason": "These are separate date-specific triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47227", "right": "issue:47256", "accept": false, - "reason": "Same title pattern, but not the same underlying report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47309", "accept": false, - "reason": "Different dated triage threads." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { - "left": "issue:47309", - "right": "issue:47419", + "left": "issue:47227", + "right": "issue:47279", "accept": false, - "reason": "Separate daily triage issues, not duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { - "left": "issue:47227", - "right": "issue:47279", + "left": "issue:47309", + "right": "issue:47419", "accept": false, - "reason": "These appear to be different days\u2019 triage reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47182", "right": "issue:47227", "accept": false, - "reason": "Adjacent daily triage issues; template similarity only." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47227", "right": "issue:47265", "accept": false, - "reason": "Different dated triage issue instances." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47419", "right": "issue:47442", "accept": false, - "reason": "These are consecutive daily triage threads, not one duplicated issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47377", "right": "issue:47419", "accept": false, - "reason": "Same series, but the date-stamped issues are distinct." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47377", "right": "issue:47442", "accept": false, - "reason": "Different daily triage threads; no mergeable duplicate." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47279", "right": "issue:47419", "accept": false, - "reason": "Distinct day-specific integration triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47406", "right": "issue:47419", "accept": false, - "reason": "Separate dated triage issues, not the same report." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47279", "right": "issue:47442", "accept": false, - "reason": "Same naming pattern, but these are different daily threads." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47227", "right": "issue:47442", "accept": false, - "reason": "Different daily triage issues with the same template." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47390", "right": "issue:47419", "accept": false, - "reason": "These are day-specific triage threads, not duplicates." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47390", "right": "issue:47406", "accept": false, - "reason": "Separate dated issue reports in the same series." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47390", "right": "issue:47442", "accept": false, - "reason": "Different daily triage threads; no evidence of one underlying duplicate." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47406", "right": "issue:47442", "accept": false, - "reason": "Consecutive daily triage issues, but not the same issue." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47377", "right": "issue:47406", "accept": false, - "reason": "Distinct date-specific triage issues." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." }, { "left": "issue:47279", "right": "issue:47406", "accept": false, - "reason": "Same triage stream, but separate daily reports." + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." } ] }, @@ -24056,7 +24105,7 @@ "issue:47377|issue:47390", "issue:47227|issue:47390" ], - "prepared_review_unit_hash": "b46511e16678980cc427a8a87b37578e40290b1f919ac6cb42c7dd8f8616efc8", + "prepared_review_unit_hash": "a0db6f763ee66bf4762f1df75605b5d5fcf01626596a6a3fea654df2baf28bf9", "component_budget": { "node_count": 23, "item_count": 23, @@ -24078,24 +24127,24 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are recurring \"[serge] integration failure triage\" issues for different dates. They look like separate daily triage reports rather than duplicates of one underlying bug, so the cluster should not be merged; if one issue must represent the set, the most referenced one is the best anchor.", - "confidence": 0.91, - "canonical_issue_reason": "issue:47377 has the strongest signal as the anchor: it is the same triage series, and it has the highest inbound reference count among the three.", + "summary": "These are three date-stamped Serge integration-failure triage issues with very similar templates, but they appear to be separate daily reports rather than duplicates of one underlying bug.", + "confidence": 0.88, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:47377 is the best single issue in the cluster because it has the most inbound references and sits in the middle of the dated series, making it the most central triage record.", + "best_issue_reason": "issue:47377 is the best representative if one is needed: it has the highest inbound references and is a central mid-series triage entry, but it should not be treated as a duplicate of the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47377", "right": "issue:47390", "accept": false, - "reason": "Reject: same template/title family, but the date suffix indicates separate daily integration-failure triage reports, not the same underlying bug." + "reason": "Same triage template and subsystem, but the titles and dates indicate separate daily integration-failure reports rather than the same underlying issue." }, { "left": "issue:47227", "right": "issue:47390", "accept": false, - "reason": "Reject: these are consecutive dated triage issues; similarity is from the shared reporting format, not evidence of one duplicate defect." + "reason": "These look like distinct dated triage notes for different days; similarity is driven by the shared reporting format, not a single bug report." } ] }, @@ -24125,7 +24174,7 @@ "soft_pairs": [ "issue:46983|issue:46984" ], - "prepared_review_unit_hash": "373d0b04a0299e52cbd0ed67ae70bc4c5052bb28810755844aae39286bef3dfd", + "prepared_review_unit_hash": "ff7662b123482fb8a86aa89464bd8d78980935f73e70df36e5383ad3fa883d58", "component_budget": { "node_count": 2, "item_count": 2, @@ -24147,27 +24196,27 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both issues share the exact same title and highly overlapping bodies, so they are likely duplicates for the same RWKV7 support request. issue:46983 is the better canonical issue because it is slightly earlier and has the fuller description.", - "confidence": 0.94, - "canonical_issue_reason": "issue:46983 is the earlier report and appears to have the more complete body, making it the best canonical issue to keep.", + "summary": "These are near-identical duplicate issues requesting RWKV7 model support; the earlier issue is the best canonical representative.", + "confidence": 0.99, + "canonical_issue_reason": "issue:46983 was opened first and has the same title and apparent request as issue:46984, making it the cleanest canonical issue.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46983 is the strongest standalone issue because it is earlier and provides the fuller problem description; issue:46984 looks like a likely duplicate rather than a distinct report.", + "best_issue_reason": "issue:46983 is the strongest single issue to keep because it is earlier and equivalent in scope to issue:46984.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46983", "right": "issue:46984", "accept": true, - "reason": "Same title and very similar body content indicate the same underlying feature request. This is a likely duplicate, though the packet does not show an explicit follow-up relationship." + "reason": "Same title and highly similar body content indicate the same underlying feature request; these are duplicates." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The duplicate judgment is mostly grounded, but the analyst is a bit too strong in places. The packet supports that the issues have the same title and highly overlapping bodies, so they are likely duplicates. However, calling issue:46984 a \"duplicate follow-up\" and describing the pair as \"near-certain\" goes beyond the evidence we have; the packet only shows similarity, not an explicit follow-up relationship. Please keep the soft-edge verdict more conservative and phrase the relationship as likely/possible duplicate rather than definitive." + "accept": true, + "feedback": "The conclusion is grounded: both issues share the same title, are close in time, and have strong soft-similarity evidence, so treating issue:46983 as the canonical duplicate representative is reasonable and conservative." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -24191,12 +24240,12 @@ "soft_pairs": [ "issue:47007|issue:47015" ], - "prepared_review_unit_hash": "e0dac138b4438b0ece1bef7e6d1cade3b80475fd7d8f6b8256a8029cb4e3a0ff", + "prepared_review_unit_hash": "fc858d2d888b0345003aa6253740706a2b6ac05229c066a8c30504b441c5546c", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1625, + "serialized_chars": 1626, "estimated_input_tokens": 407, "estimated_eval_tokens": 1070 }, @@ -24204,7 +24253,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1625, + "serialized_chars": 1626, "estimated_input_tokens": 407, "estimated_eval_tokens": 1070 }, @@ -24213,18 +24262,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are high-level writeups about AI/LLM-driven vulnerability discovery, but they read like different topic proposals rather than the same underlying issue. I would not merge them as duplicates.", - "confidence": 0.82, - "canonical_issue_reason": "Issue 47015 is the broader umbrella topic; 47007 is a more specific \u201clessons from using advanced LLMs\u201d variant.", + "summary": "Both items read like standalone article/discussion posts about AI-assisted vulnerability discovery, but they describe different framing and do not point to the same concrete bug, feature, or task. They should not be deduplicated.", + "confidence": 0.93, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "47015 is the best representative of the cluster because it has the clearest, most general framing of the shared AI-vulnerability-discovery theme.", + "best_issue_reason": "Issue 47015 is the broader, more general title and is the better representative if this cluster is being summarized, but it is not a duplicate of 47007.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47007", "right": "issue:47015", "accept": false, - "reason": "Same general subject area, but not the same concrete bug/change. The titles suggest different angles and likely different writeups, not one underlying duplicate issue." + "reason": "The titles and bodies suggest related topic area only; they do not identify the same underlying bug/change or a single concrete target, so this is not a duplicate." } ] }, @@ -24254,7 +24303,7 @@ "soft_pairs": [ "issue:47138|issue:47139" ], - "prepared_review_unit_hash": "36b8688902d81b16a6e11706805b1051e266fdef5251e43ec736976008b19a87", + "prepared_review_unit_hash": "b92daa53f542c8f2543fc8e13264d525c76f1cbe4a891b68e53f223f0b4cc3ad", "component_budget": { "node_count": 2, "item_count": 2, @@ -24276,24 +24325,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two items are near-perfect duplicates: identical titles, matching body size, and the same reported bug about chat routes management endpoints under /v1 when base_url includes /v1.", + "summary": "These are near-certain duplicate issue reports with the same title and effectively identical content; keep the slightly earlier, more-linked issue as canonical.", "confidence": 0.99, - "canonical_issue_reason": "Issue #47138 is the better canonical candidate because it was created first and has slightly more discussion/inbound references, suggesting marginally higher centrality.", + "canonical_issue_reason": "Issue #47138 is earlier and has more inbound references, making it the better canonical thread for the duplicate report.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #47138 is the best global issue representative for the same reasons: earliest creation and slightly stronger apparent traction.", + "best_issue_reason": "Issue #47138 is the strongest representative because it is first-created and has the higher inbound reference count while describing the same bug as #47139.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:47138", "right": "issue:47139", "accept": true, - "reason": "Same title and same underlying bug report; no distinguishing evidence suggests different problems." + "reason": "Same title and same reported bug; these are duplicates of the same issue, so they should be clustered together." } ] }, "evaluator_result": { "accept": true, - "feedback": "The assessment is grounded in the packet: the issues have identical titles, identical body length, and near-identical metadata, with 47138 slightly ahead on created_at, discussion activity, and inbound references. The soft-edge accept is conservative enough, and no unsupported claims are made." + "feedback": "Grounded and conservative: the two issues have identical titles, near-simultaneous creation times, and strong soft-similarity evidence, while the canonical choice is reasonably supported by earlier creation and higher inbound references." }, "evaluator_used": true, "retried": false, @@ -24312,14 +24361,14 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-96", + "cluster_id": "cluster-40546-97", "nodes": [ "pull_request:43251", "pull_request:43254", "pull_request:44775", - "pull_request:46196", "pull_request:46266", "pull_request:46381", + "pull_request:46836", "pull_request:46912", "pull_request:46957", "pull_request:47077", @@ -24332,7 +24381,6 @@ "pull_request:47129", "pull_request:47166", "pull_request:47173", - "pull_request:47179", "pull_request:47202", "pull_request:47204", "pull_request:47212", @@ -24361,7 +24409,8 @@ "pull_request:47447", "pull_request:47449", "pull_request:47450", - "pull_request:47457" + "pull_request:47457", + "pull_request:47462" ], "soft_pairs": [ "pull_request:43251|pull_request:43254", @@ -24371,8 +24420,8 @@ "pull_request:47282|pull_request:47444", "pull_request:47123|pull_request:47281", "pull_request:47124|pull_request:47380", - "pull_request:47122|pull_request:47280", "pull_request:47122|pull_request:47224", + "pull_request:47122|pull_request:47280", "pull_request:47124|pull_request:47445", "pull_request:47310|pull_request:47445", "pull_request:47282|pull_request:47445", @@ -24402,8 +24451,10 @@ "pull_request:47360|pull_request:47447", "pull_request:47384|pull_request:47447", "pull_request:44775|pull_request:47275", + "pull_request:47301|pull_request:47462", "pull_request:46957|pull_request:47216", "pull_request:46381|pull_request:47457", + "pull_request:46836|pull_request:47462", "pull_request:47166|pull_request:47173", "pull_request:47291|pull_request:47384", "pull_request:47216|pull_request:47275", @@ -24412,302 +24463,288 @@ "pull_request:47294|pull_request:47450", "pull_request:47275|pull_request:47301", "pull_request:47216|pull_request:47384", - "pull_request:46912|pull_request:47284", - "pull_request:46912|pull_request:47179", - "pull_request:46196|pull_request:47202" + "pull_request:46912|pull_request:47284" ], - "prepared_review_unit_hash": "d92b12119b09e8fed836d8e857fa1678826145f843b7c7b00a1b58085cdb5676", + "prepared_review_unit_hash": "e5ec8d301ea0629a53658f18edb19f26125d2c51f3a508e1bff44896e487d043", "component_budget": { - "node_count": 96, - "item_count": 96, - "soft_pair_count": 272, - "serialized_chars": 334252, - "estimated_input_tokens": 83563, - "estimated_eval_tokens": 167382 + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 51, - "serialized_chars": 127735, - "estimated_input_tokens": 31934, - "estimated_eval_tokens": 64124 + "serialized_chars": 126888, + "estimated_input_tokens": 31722, + "estimated_eval_tokens": 63700 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous cluster with a few true near-duplicate pairs: one cross-entropy kwargs fix and a few same-model test-failure remediations. Most other soft links are just broad topical similarity and should stay separate.", - "confidence": 0.81, + "summary": "This cluster is mostly heterogeneous and does not form a single duplicate set. The only clear duplicate pair is the two fixed_cross_entropy PRs; one other edge (EdgeTam original_sizes handling) is plausibly the same bug, but most remaining soft links are false positives across unrelated models, docs, CI, exporter, and quantization changes.", + "confidence": 0.96, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "43251 is the cleanest representative PR in the set: it has an explicit issue link, a narrow single-function fix, and it neatly subsumes the companion kwargs variant 43254.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "Same file/function, same issue 43240, and nearly identical cross_entropy argument plumbing; these look like alternate fixes for the same change." + "reason": "Same underlying fix in src/transformers/loss/loss_utils.py: forward supported kwargs to nn.functional.cross_entropy, including weight and label_smoothing; both target issue 43240." }, { "left": "pull_request:47282", "right": "pull_request:47310", "accept": false, - "reason": "Both are test-fix PRs, but they target different models and different failures (Mamba2 OOM vs Gemma3 output mismatch)." + "reason": "Different model/test failures and different code paths: Mamba2 OOM workaround vs Gemma3 output mismatch." }, { "left": "pull_request:47124", "right": "pull_request:47282", "accept": false, - "reason": "Different models and failure types; these are unrelated test adjustments, not the same bug." + "reason": "Both are test-fix PRs, but they address different models and different failure mechanics (Falcon-Mamba output mismatch vs Mamba2 OOM)." }, { "left": "pull_request:47280", "right": "pull_request:47282", "accept": false, - "reason": "Same broad serge-style pattern, but different model/test files and different failures." + "reason": "Same model family, but not the same concrete change: one adds temp offload handling for loading, the other changes generation length/cache cleanup for OOM." }, { "left": "pull_request:47282", "right": "pull_request:47444", "accept": false, - "reason": "Unrelated models and code paths; the DeepSeek/MPT changes do not match the Mamba2 OOM fix." + "reason": "Unrelated failures in different models: Mamba2 OOM vs MPT load_error." }, { "left": "pull_request:47123", "right": "pull_request:47281", "accept": false, - "reason": "Different model families and different fix targets; no shared concrete bug." + "reason": "Different model integrations and failures: MiniMax-M3-VL load_error vs GLM46V test failures." }, { "left": "pull_request:47124", "right": "pull_request:47380", "accept": false, - "reason": "Different model failures (falcon_mamba vs mamba2), so these are not duplicates." + "reason": "Both touch test stability, but they fix different models and different symptoms; not the same underlying change." }, { "left": "pull_request:47122", - "right": "pull_request:47280", - "accept": false, - "reason": "Different model/test targets; the Edgetam fix and MiniMax-M3-VL fix are unrelated." + "right": "pull_request:47224", + "accept": true, + "reason": "Same EdgeTam bug: both drop original_sizes before calling the TimmWrapper backbone because the backbone does not accept that kwarg." }, { "left": "pull_request:47122", - "right": "pull_request:47224", - "accept": true, - "reason": "Both address Edgetam passing `original_sizes` into a backbone that does not accept it; 47224 is broader, but the Edgetam fix is the same underlying bug." + "right": "pull_request:47280", + "accept": false, + "reason": "EdgeTam import_or_config failure is unrelated to MiniMax-M3-VL loading/OOM handling." }, { "left": "pull_request:47124", "right": "pull_request:47445", "accept": false, - "reason": "Different models and different failure modes; not the same underlying issue." + "reason": "Falcon-Mamba output mismatch and PEFT/DeepSeek_v3 loading issues are unrelated." }, { "left": "pull_request:47310", "right": "pull_request:47445", "accept": false, - "reason": "Gemma3 generation output mismatch and PEFT/DeepSeek integration failures are unrelated." + "reason": "Gemma3 generation mismatch vs PEFT/DeepSeek_v3 loading behavior; no shared bug or code path." }, { "left": "pull_request:47282", "right": "pull_request:47445", "accept": false, - "reason": "No shared code path or model; these are separate integration fixes." + "reason": "Different models and failures; no common concrete fix." }, { "left": "pull_request:47124", "right": "pull_request:47443", "accept": false, - "reason": "Falcon-Mamba output mismatch and DeepSeek-V32 load-related cleanup are unrelated." + "reason": "Different model families and failure modes; not a duplicate." }, { "left": "pull_request:47281", "right": "pull_request:47445", "accept": false, - "reason": "Different model families and different failure classes; not mergeable as one PR." + "reason": "GLM46V integration failures and PEFT/DeepSeek_v3 loading issues are unrelated." }, { "left": "pull_request:47268", "right": "pull_request:47284", "accept": false, - "reason": "Different models and different symptoms; one is Zaya test cleanup, the other is MIMO V2 Flash model fix." + "reason": "Different test failures in different models (Zaya vs MIMO v2 Flash)." }, { "left": "pull_request:47173", "right": "pull_request:47284", "accept": false, - "reason": "XPU expectation updates for different test suites are not the same change." + "reason": "XPU expectation updates across multiple models are unrelated to MIMO v2 Flash test fixes." }, { "left": "pull_request:47284", "right": "pull_request:47449", "accept": false, - "reason": "Different models and different code paths; these are unrelated test/model fixes." + "reason": "MIMO v2 Flash test fix and Qwen3 Omni Moe test fix are separate model-specific changes." }, { "left": "pull_request:47202", "right": "pull_request:47204", "accept": false, - "reason": "Both touch docker/CI infrastructure, but one removes a workflow while the other adds exporter test deps; not the same bug/change." + "reason": "Different concerns: deleting CI docker-image builder workflows vs adding exporter test dependencies to torch-light." }, { "left": "pull_request:47204", "right": "pull_request:47374", "accept": false, - "reason": "A docker image dependency update and exporter documentation are separate concerns." + "reason": "Exporter-test environment setup is unrelated to exporter docs restructuring." }, { "left": "pull_request:47284", "right": "pull_request:47292", "accept": false, - "reason": "Different model/test suites with different failures; no concrete overlap." + "reason": "Different models and failure modes; no shared fix." }, { "left": "pull_request:47108", "right": "pull_request:47202", "accept": false, - "reason": "Doc-builder comment cleanup and docker workflow deletion are not the same underlying change." + "reason": "Doc-builder comment cleanup and docker-image builder removal are unrelated changes." }, { "left": "pull_request:47077", "right": "pull_request:47078", "accept": false, - "reason": "Both are Japanese docs PRs, but they make different wording/content edits and are not duplicates." + "reason": "Both edit Japanese docs, but one is a DistilBERT translation addition while the other is broader wording cleanup; overlap in toctree is incidental." }, { "left": "pull_request:47224", "right": "pull_request:47457", "accept": false, - "reason": "Edgetam/Whisper/tokenizer fixes and Inkling auto-fix are unrelated." - }, - { - "left": "pull_request:47105", - "right": "pull_request:47129", - "accept": false, - "reason": "VITS test re-enable and RT-DETR converter updates are unrelated." + "reason": "EdgeTam import_or_config fix and Inkling test adjustments are unrelated." }, { - "left": "pull_request:47449", - "right": "pull_request:47457", + "left": "pull_request:47204", + "right": "pull_request:47243", "accept": false, - "reason": "Different model code paths; these do not fix the same bug." + "reason": "Both concern exporters, but one is dependency/test-environment setup and the other is exporter API/behavior changes; not the same code-path fix." }, { - "left": "pull_request:46266", - "right": "pull_request:46957", + "left": "pull_request:47202", + "right": "pull_request:47360", "accept": false, - "reason": "Both are large PRs with model/integration work, but they touch different subsystems and changesets." + "reason": "Docker builder removal and OTEL CI workflow changes are distinct infrastructure changes." }, { - "left": "pull_request:47077", - "right": "pull_request:47275", + "left": "pull_request:47108", + "right": "pull_request:47360", "accept": false, - "reason": "Japanese docs wording changes are unrelated to EfficientViT-SAM model support." + "reason": "Dependabot-friendly doc-builder comments are unrelated to the OTEL CI workflow trim." }, { - "left": "pull_request:47294", - "right": "pull_request:47450", + "left": "pull_request:47292", + "right": "pull_request:47457", "accept": false, - "reason": "Kernel-native causal_conv1d dispatch and linear-attention dispatch simplification are different changes." + "reason": "Different models and failure types; no duplicate relationship." }, { - "left": "pull_request:47275", - "right": "pull_request:47301", + "left": "pull_request:47441", + "right": "pull_request:47457", "accept": false, - "reason": "EfficientViT-SAM model support and MTP docs are unrelated." + "reason": "Whisper dtype/max_length test fixes and Inkling test fixes are unrelated." }, { - "left": "pull_request:47216", - "right": "pull_request:47275", + "left": "pull_request:47105", + "right": "pull_request:47129", "accept": false, - "reason": "Compressed-tensors fp8 and EfficientViT-SAM are distinct features with no shared code path." + "reason": "VITS test re-enable and RT-DETR v2 converter fixes are unrelated." }, { - "left": "pull_request:47216", + "left": "pull_request:47291", "right": "pull_request:47384", "accept": false, - "reason": "Quantization docs/code and Gemma 4 heterogeneous attention config are unrelated." + "reason": "Deprecating continuous_batching_config and explicit Gemma 4 heterogeneous config handling are separate generation-config changes." }, { - "left": "pull_request:47166", - "right": "pull_request:47173", + "left": "pull_request:47216", + "right": "pull_request:47275", "accept": false, - "reason": "GPTQ/MXFP4 quantization tests and XPU expectation updates are different test fixes." + "reason": "Compressed-tensors fp8 support and EfficientViT-SAM model addition are unrelated." }, { - "left": "pull_request:47291", - "right": "pull_request:47384", + "left": "pull_request:47077", + "right": "pull_request:47275", "accept": false, - "reason": "Continuous batching deprecation and Gemma 4 heterogeneous config are separate generation/config changes." + "reason": "Japanese DistilBERT docs and EfficientViT-SAM model addition are unrelated." }, { - "left": "pull_request:47216", + "left": "pull_request:47294", "right": "pull_request:47450", "accept": false, - "reason": "Compressed-tensors fp8 and linear-attention dispatch simplification are unrelated changes." - }, - { - "left": "pull_request:47301", - "right": "pull_request:47360", - "accept": false, - "reason": "MTP docs and the OTEL CI workflow pilot are unrelated." + "reason": "Both touch integration/kernel dispatch helpers, but they implement different features for different model paths and are not the same fix." }, { - "left": "pull_request:47360", - "right": "pull_request:47450", + "left": "pull_request:47275", + "right": "pull_request:47301", "accept": false, - "reason": "OTEL CI workflow changes and linear-attention dispatch are unrelated." + "reason": "EfficientViT-SAM model addition and MTP docs are unrelated." }, { - "left": "pull_request:47360", - "right": "pull_request:47447", + "left": "pull_request:46957", + "right": "pull_request:47216", "accept": false, - "reason": "OTEL CI workflow changes and speculative decoding cache rollback are unrelated." + "reason": "Finegrained FP8 generation/quantizer changes and compressed-tensors fp8 are different quantization features." }, { - "left": "pull_request:47384", - "right": "pull_request:47447", + "left": "pull_request:46381", + "right": "pull_request:47457", "accept": false, - "reason": "Gemma 4 heterogeneous config and speculative-decoding cache rollback solve different problems." + "reason": "Tokenizers version pinning and Inkling test fixes are unrelated." }, { - "left": "pull_request:44775", - "right": "pull_request:47275", + "left": "pull_request:46836", + "right": "pull_request:47462", "accept": false, - "reason": "Japanese N-D parallelism docs and EfficientViT-SAM model support are unrelated." + "reason": "Both touch conversion_mapping, but one is Dinov2/DINO-related model conversion and the other is MTP/GLM OCR mapping; different code paths." }, { - "left": "pull_request:46957", - "right": "pull_request:47216", + "left": "pull_request:47166", + "right": "pull_request:47173", "accept": false, - "reason": "Both are quantization-adjacent, but they target different mechanisms and code paths." + "reason": "GPTQ quantization tests and XPU expectation updates are unrelated." }, { - "left": "pull_request:46381", - "right": "pull_request:47457", + "left": "pull_request:47291", + "right": "pull_request:47384", "accept": false, - "reason": "Tokenizer version compatibility and Inkling test/autofix changes are unrelated." + "reason": "Continuous-batching deprecation and Gemma 4 heterogeneous config changes are different configuration changes." }, { - "left": "pull_request:46912", - "right": "pull_request:47284", + "left": "pull_request:47216", + "right": "pull_request:47450", "accept": false, - "reason": "RoPE test improvements and MIMO V2 Flash fixes are different changes." + "reason": "Compressed-tensors support and linear-attention kernel dispatch simplification are unrelated." }, { - "left": "pull_request:46912", - "right": "pull_request:47179", + "left": "pull_request:47077", + "right": "pull_request:47275", "accept": false, - "reason": "RoPE test improvements and Granite-SWA model support are unrelated." + "reason": "Doc translation work and new model addition are unrelated." }, { - "left": "pull_request:46196", - "right": "pull_request:47202", + "left": "pull_request:47294", + "right": "pull_request:47450", "accept": false, - "reason": "SDPA dispatch support and docker image builder removal are separate subsystems with no shared bug." + "reason": "Kernel-native causal_conv1d support and linear-attention dispatch simplification are different features, not one duplicate change." } ] }, @@ -24729,7 +24766,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-96", + "cluster_id": "cluster-40546-97", "nodes": [ "pull_request:43451", "pull_request:44517", @@ -24739,7 +24776,6 @@ "pull_request:46266", "pull_request:46395", "pull_request:46509", - "pull_request:46556", "pull_request:46604", "pull_request:46658", "pull_request:46836", @@ -24778,25 +24814,32 @@ "pull_request:47450", "pull_request:47451", "pull_request:47452", - "pull_request:47461" + "pull_request:47461", + "pull_request:47462" ], "soft_pairs": [ + "pull_request:46912|pull_request:47179", + "pull_request:46196|pull_request:47202", "pull_request:47204|pull_request:47357", "pull_request:46196|pull_request:47210", - "pull_request:47003|pull_request:47204", "pull_request:47166|pull_request:47449", + "pull_request:47003|pull_request:47204", "pull_request:46836|pull_request:47202", "pull_request:47093|pull_request:47275", + "pull_request:47082|pull_request:47462", "pull_request:45157|pull_request:47093", "pull_request:46995|pull_request:47163", "pull_request:45157|pull_request:46395", "pull_request:46836|pull_request:47275", + "pull_request:46266|pull_request:47462", "pull_request:47204|pull_request:47295", + "pull_request:47216|pull_request:47462", "pull_request:46912|pull_request:47082", "pull_request:47163|pull_request:47384", "pull_request:44775|pull_request:47301", - "pull_request:47301|pull_request:47447", "pull_request:46196|pull_request:46995", + "pull_request:47301|pull_request:47447", + "pull_request:47447|pull_request:47462", "pull_request:47447|pull_request:47450", "pull_request:47301|pull_request:47384", "pull_request:47243|pull_request:47304", @@ -24810,23 +24853,23 @@ "pull_request:47294|pull_request:47320", "pull_request:47113|pull_request:47422", "pull_request:47163|pull_request:47357", + "pull_request:46196|pull_request:47179", "pull_request:47093|pull_request:47143", "pull_request:46509|pull_request:46995", - "pull_request:46196|pull_request:47179", "pull_request:46995|pull_request:47143", "pull_request:46266|pull_request:47301", + "pull_request:47450|pull_request:47462", "pull_request:44775|pull_request:47304", "pull_request:47093|pull_request:47163", "pull_request:47082|pull_request:47210", - "pull_request:46509|pull_request:47357", "pull_request:47179|pull_request:47210", + "pull_request:46509|pull_request:47357", "pull_request:46604|pull_request:47003", "pull_request:47082|pull_request:47108", + "pull_request:47078|pull_request:47275", "pull_request:47108|pull_request:47301", "pull_request:43451|pull_request:47357", - "pull_request:47078|pull_request:47275", "pull_request:46266|pull_request:47216", - "pull_request:47082|pull_request:47275", "pull_request:46266|pull_request:47336", "pull_request:46995|pull_request:47301", "pull_request:47175|pull_request:47179", @@ -24835,363 +24878,368 @@ "pull_request:47163|pull_request:47179", "pull_request:47320|pull_request:47452", "pull_request:44517|pull_request:45157", - "pull_request:46266|pull_request:47452", - "pull_request:47210|pull_request:47304", "pull_request:47275|pull_request:47336", - "pull_request:46556|pull_request:46836", - "pull_request:47301|pull_request:47450", + "pull_request:47210|pull_request:47304", "pull_request:46196|pull_request:47461", "pull_request:47204|pull_request:47422", "pull_request:47113|pull_request:47452", "pull_request:47336|pull_request:47357", - "pull_request:46196|pull_request:47450" + "pull_request:46912|pull_request:47078", + "pull_request:46196|pull_request:47450", + "pull_request:47108|pull_request:47204" ], - "prepared_review_unit_hash": "14909caa413c82ca06059a636617576d50e950a12013bf1fcd0fe091566d8f9a", + "prepared_review_unit_hash": "084c3cca603af894aca73cfe5ca1f3b6e2e05d18f09a83af15f88fd2d1ef87e3", "component_budget": { - "node_count": 96, - "item_count": 96, - "soft_pair_count": 272, - "serialized_chars": 334252, - "estimated_input_tokens": 83563, - "estimated_eval_tokens": 167382 + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 64, - "serialized_chars": 140805, - "estimated_input_tokens": 35202, - "estimated_eval_tokens": 70660 + "soft_pair_count": 69, + "serialized_chars": 141333, + "estimated_input_tokens": 35334, + "estimated_eval_tokens": 70924 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of unrelated PRs: new model integrations, documentation updates, backend/runtime support, CI/test fixes, and a few isolated bugfixes. The soft edges are mostly driven by shared subsystems or filenames, not by the same underlying change, so none should be merged as duplicates.", - "confidence": 0.98, + "summary": "The cluster is highly heterogeneous: it mixes unrelated model additions, docs-only updates, CI/docker changes, exporter work, attention/backend fixes, and test tweaks. None of the soft pairs looks like the same underlying bug or change, so I would not collapse any of them as duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single PR is a safe canonical representative; the items span unrelated features and fixes across models, docs, CI, and export/runtime plumbing.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "No global best PR for deduping: the cluster is not one coherent change, so picking any PR would be arbitrary.", "soft_edge_verdicts": [ + { + "left": "pull_request:46912", + "right": "pull_request:47179", + "accept": false, + "reason": "Both are RoPE-related only in the broadest sense, but one is test coverage and the other is new Granite SWA model support; different changes." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47202", + "accept": false, + "reason": "One is SDPA dispatch for audio models, the other is removing a CI docker builder; unrelated code paths." + }, { "left": "pull_request:47204", "right": "pull_request:47357", "accept": false, - "reason": "Both touch CI/runtime packaging, but one adds exporter test deps to torch-light while the other adds FSDP CI and save support; different problems and merge scopes." + "reason": "Exporter test-image dependency changes versus FSDP CI/save logic; not the same bug or feature." }, { "left": "pull_request:46196", "right": "pull_request:47210", "accept": false, - "reason": "Both modify sdpa attention code, but one is broad audio-model SDPA dispatch support and the other is a specific packed-row varlen_attn routing change." + "reason": "Both touch attention internals, but one is model dispatch support and the other is packed-row varlen attention; different fixes." }, { - "left": "pull_request:47003", - "right": "pull_request:47204", + "left": "pull_request:47166", + "right": "pull_request:47449", "accept": false, - "reason": "OpenVINO exporter support and torch-light exporter test dependencies are separate features with different code paths and goals." + "reason": "Quantization test updates versus qwen3_omni_moe model test fixes; unrelated." }, { - "left": "pull_request:47166", - "right": "pull_request:47449", + "left": "pull_request:47003", + "right": "pull_request:47204", "accept": false, - "reason": "Quantization test adjustments and Qwen3 Omni-MoE test fixes are unrelated model/backend issues." + "reason": "Both are exporter/tooling related, but one adds OpenVINO exporter support and the other changes CI image dependencies." }, { "left": "pull_request:46836", "right": "pull_request:47202", "accept": false, - "reason": "A new OCR model integration is unrelated to removing a CI docker image builder." + "reason": "Unlimited OCR model support versus docker image builder removal; unrelated." }, { "left": "pull_request:47093", "right": "pull_request:47275", "accept": false, - "reason": "Both add models, but they are different architectures with distinct configs, processors, and model code." + "reason": "Different new models, different architectures, no shared concrete change." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47462", + "accept": false, + "reason": "RoMa keypoint matcher support versus MTP/GLM OCR generation changes; unrelated." }, { "left": "pull_request:45157", "right": "pull_request:47093", "accept": false, - "reason": "PrismML Bonsai support and OpenPangu-V2 support are unrelated model integrations." + "reason": "PrismML GGUF support versus OpenPangu-V2 model addition; different subsystems." }, { "left": "pull_request:46995", "right": "pull_request:47163", "accept": false, - "reason": "Backend availability for flash-linear-attention on MLU and XPU paged-attention test coverage are different changes." + "reason": "A backend availability helper change versus XPU test additions; not the same bug." }, { "left": "pull_request:45157", "right": "pull_request:46395", "accept": false, - "reason": "Different model additions: PrismML Bonsai support versus Param2MoE architecture support." + "reason": "Prism GGUF support versus Param2MoE model support; unrelated." }, { "left": "pull_request:46836", "right": "pull_request:47275", "accept": false, - "reason": "Unlimited OCR and EfficientViT-SAM are unrelated model additions." + "reason": "Two different model additions with no shared code path." }, { - "left": "pull_request:47204", - "right": "pull_request:47295", + "left": "pull_request:46266", + "right": "pull_request:47462", "accept": false, - "reason": "Exporter test dependency changes do not overlap with DiffusionGemma mask-casting bugfixes." + "reason": "DINO conversion/modernization work versus MTP generation support; unrelated changes." }, { "left": "pull_request:46912", "right": "pull_request:47082", "accept": false, - "reason": "RoPE test coverage and RoMa keypoint matcher are unrelated; shared 'model support' only at a very broad level." + "reason": "RoPE test coverage versus a keypoint matcher model addition; not duplicate work." }, { "left": "pull_request:47163", "right": "pull_request:47384", "accept": false, - "reason": "XPU paged-attention tests and Gemma 4 heterogeneous config changes are unrelated." + "reason": "Paged-attention XPU tests versus Gemma4 heterogeneity config changes; different issues." }, { "left": "pull_request:44775", "right": "pull_request:47301", "accept": false, - "reason": "Both are docs-only, but one is parallelism docs and the other is assisted decoding/MTP docs; different topics." + "reason": "Docs navigation/parallelism updates versus assisted-decoding MTP docs; different documentation topics." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "Audio SDPA dispatch versus flash-linear-attention availability on MLU; different backend problems." }, { "left": "pull_request:47301", "right": "pull_request:47447", "accept": false, - "reason": "Assisted-decoding docs for MTP and cache rollback for speculative decoding are related to generation, but they fix different mechanisms." + "reason": "Assisted decoding docs versus cache rollback behavior for speculative decoding; not the same change." }, { - "left": "pull_request:46196", - "right": "pull_request:46995", + "left": "pull_request:47301", + "right": "pull_request:47384", "accept": false, - "reason": "Both concern attention/backends, but one is SDPA dispatch for audio models and the other is flash-linear-attention availability on MLU." + "reason": "MTP documentation versus Gemma4 heterogeneous config implementation; unrelated." + }, + { + "left": "pull_request:47447", + "right": "pull_request:47462", + "accept": false, + "reason": "Cache rollback for speculative decoding versus MTP for GLM OCR; different generation mechanisms." }, { "left": "pull_request:47447", "right": "pull_request:47450", "accept": false, - "reason": "Speculative-decoding cache rollback and linear-attention dispatch simplification are different code paths and problems." + "reason": "Speculative-decoding cache changes versus linear-attention dispatch simplification; unrelated." }, { "left": "pull_request:47301", - "right": "pull_request:47384", + "right": "pull_request:47422", "accept": false, - "reason": "MTP generation docs and Gemma 4 heterogeneous-attention config changes are unrelated." + "reason": "MTP docs versus CI/docker pinning and RNNT fixes; not duplicate." }, { "left": "pull_request:47243", "right": "pull_request:47304", "accept": false, - "reason": "Executorch exporter fixes and an Apple Silicon docs cleanup are unrelated." + "reason": "Executorch exporter fixes versus MPS graph-cache docs cleanup; unrelated." }, { "left": "pull_request:47077", "right": "pull_request:47301", "accept": false, - "reason": "Japanese DistilBERT docs translation and MTP docs are unrelated documentation changes." + "reason": "Japanese docs translation updates versus English MTP docs; different documentation work." }, { "left": "pull_request:47243", "right": "pull_request:47447", "accept": false, - "reason": "Exporter fixes and speculative-decoding cache rollback are separate runtime features." + "reason": "Executorch exporter fixes versus cache rollback support; different code paths." }, { "left": "pull_request:46604", "right": "pull_request:47357", "accept": false, - "reason": "Mistral tekken tokenizer utilities and FSDP CI/save changes are unrelated." + "reason": "Mistral tekken tokenizer utilities versus FSDP CI/save changes; unrelated." }, { "left": "pull_request:47163", "right": "pull_request:47451", "accept": false, - "reason": "XPU paged-attention tests and flash-attention value padding interfaces solve different issues." + "reason": "XPU paged-attention tests versus attention-interface value padding; different problems." }, { "left": "pull_request:44775", "right": "pull_request:47077", "accept": false, - "reason": "Both are documentation PRs, but they target different language/topic areas and are not the same change." + "reason": "English distributed-training docs versus Japanese DistilBERT docs translation; unrelated." }, { "left": "pull_request:47082", "right": "pull_request:47216", "accept": false, - "reason": "RoMa keypoint matcher and compressed-tensors fp8 support are unrelated model/runtime features." + "reason": "RoMa model support versus compressed-tensors FP8 quantization support; different features." }, { "left": "pull_request:47166", "right": "pull_request:47292", "accept": false, - "reason": "Quantization test tweaks and XPU test fixes for tipsv2_dpt are unrelated." + "reason": "Quantization test edits versus TipsV2 DPT XPU test fixes; unrelated." }, { "left": "pull_request:47294", "right": "pull_request:47320", "accept": false, - "reason": "Kernel native causal_conv1d functions and response_schema deprecation touch different APIs and concerns." + "reason": "Kernel-native causal-conv work versus response-schema deprecation; different areas." }, { "left": "pull_request:47113", "right": "pull_request:47422", "accept": false, - "reason": "Audio feature/process requirements fixes and torchaudio pinning/RNNT cleanup are related to audio tooling but not the same bug/change." + "reason": "Audio feature/proc requirement fixes versus CI image torch pinning and RNNT break fix; unrelated." }, { - "left": "pull_request:47163", - "right": "pull_request:47357", + "left": "pull_request:47320", + "right": "pull_request:47452", "accept": false, - "reason": "Paged-attention XPU tests and FSDP CI/save support are unrelated test infra changes." + "reason": "Response-schema deprecation versus LFM2/Mamba-style model alignment work; not the same change." }, { - "left": "pull_request:47093", - "right": "pull_request:47143", + "left": "pull_request:44517", + "right": "pull_request:45157", "accept": false, - "reason": "Two separate model integrations: OpenPangu-V2 and NeuCodec." + "reason": "Qwen3-TTS model support versus Prism GGUF dequantization support; unrelated." }, { - "left": "pull_request:46509", - "right": "pull_request:46995", + "left": "pull_request:47275", + "right": "pull_request:47336", "accept": false, - "reason": "Nemotron Omni support and flash-linear-attention backend availability are unrelated." + "reason": "EfficientViT-SAM model support versus offloaded model saving bugfix; unrelated." }, { - "left": "pull_request:46196", - "right": "pull_request:47179", + "left": "pull_request:47210", + "right": "pull_request:47304", "accept": false, - "reason": "Audio SDPA dispatch support and Granite model support are unrelated." + "reason": "Packed SDPA runtime change versus MPS graph-cache docs cleanup; unrelated." }, { - "left": "pull_request:46995", - "right": "pull_request:47143", + "left": "pull_request:46196", + "right": "pull_request:47461", "accept": false, - "reason": "MLU flash-linear-attention availability and NeuCodec support are unrelated backend/model additions." + "reason": "Audio SDPA dispatch versus Wav2Vec2Phoneme tokenizer language-state fix; different code paths." }, { - "left": "pull_request:46266", - "right": "pull_request:47301", + "left": "pull_request:47204", + "right": "pull_request:47422", "accept": false, - "reason": "DINOv2 modernization/conversion mapping and MTP docs are unrelated." + "reason": "CI/exporter dependencies versus torch pinning in quantization docker; related to CI, but not the same concrete issue." }, { - "left": "pull_request:44775", - "right": "pull_request:47304", + "left": "pull_request:47336", + "right": "pull_request:47357", "accept": false, - "reason": "Parallelism docs and MPS graph cache docs are different documentation topics." + "reason": "Fully offloaded save bugfix versus FSDP CI/end-to-end tests; different problems." }, { - "left": "pull_request:47093", - "right": "pull_request:47163", + "left": "pull_request:46912", + "right": "pull_request:47078", "accept": false, - "reason": "Model addition and XPU paged-attention test coverage are unrelated." + "reason": "RoPE testing updates versus Japanese documentation wording improvements; unrelated." }, { - "left": "pull_request:47082", - "right": "pull_request:47210", + "left": "pull_request:46196", + "right": "pull_request:47450", "accept": false, - "reason": "RoMa model support and sdpa varlen routing are unrelated." + "reason": "Audio SDPA dispatch versus linear-attention function dispatch simplification; different attention backends and code paths." }, { - "left": "pull_request:46509", - "right": "pull_request:47357", + "left": "pull_request:47093", + "right": "pull_request:47163", "accept": false, - "reason": "Nemotron Omni support and FSDP CI/save support are unrelated." + "reason": "OpenPangu-V2 model support versus XPU paged-attention tests; unrelated." }, { - "left": "pull_request:47179", + "left": "pull_request:47082", "right": "pull_request:47210", "accept": false, - "reason": "Granite SWA model support and packed-row SDPA routing are unrelated." + "reason": "RoMa model addition versus packed SDPA varlen attention; unrelated." }, { - "left": "pull_request:46604", - "right": "pull_request:47003", - "accept": false, - "reason": "Mistral tekken utilities and OpenVINO exporter support are unrelated." - }, - { - "left": "pull_request:47163", - "right": "pull_request:47179", - "accept": false, - "reason": "XPU attention tests and Granite SWA model support are unrelated." - }, - { - "left": "pull_request:47320", - "right": "pull_request:47452", + "left": "pull_request:47179", + "right": "pull_request:47210", "accept": false, - "reason": "Response-schema deprecation and mamba standardization are unrelated changes." + "reason": "Granite SWA model support versus packed SDPA runtime change; not duplicate work." }, { - "left": "pull_request:44517", - "right": "pull_request:45157", + "left": "pull_request:46509", + "right": "pull_request:47357", "accept": false, - "reason": "Qwen3-TTS support and PrismML Bonsai support are distinct model additions." + "reason": "Nemotron Omni model support versus FSDP CI/save changes; unrelated." }, { - "left": "pull_request:46266", - "right": "pull_request:47452", + "left": "pull_request:46604", + "right": "pull_request:47003", "accept": false, - "reason": "DINOv2/DINOv3 conversion modernization and mamba code cleanup are unrelated." + "reason": "Mistral tokenizer conversion utilities versus OpenVINO exporter support; both tooling-related but different concrete features." }, { - "left": "pull_request:47210", - "right": "pull_request:47304", + "left": "pull_request:47082", + "right": "pull_request:47108", "accept": false, - "reason": "Packed SDPA routing and MPS graph cache docs are unrelated." + "reason": "Model support versus doc-builder comment tweak; unrelated." }, { - "left": "pull_request:47275", - "right": "pull_request:47336", + "left": "pull_request:47078", + "right": "pull_request:47275", "accept": false, - "reason": "EfficientViT-SAM model addition and fully offloaded model saving bugfix are unrelated." + "reason": "Japanese docs wording changes versus EfficientViT-SAM model support; unrelated." }, { - "left": "pull_request:46556", - "right": "pull_request:46836", + "left": "pull_request:47163", + "right": "pull_request:47357", "accept": false, - "reason": "Processor/image utility updates and Unlimited OCR model support are not the same change." + "reason": "XPU paged-attention tests versus FSDP CI/save work; different issues." }, { - "left": "pull_request:47301", - "right": "pull_request:47450", + "left": "pull_request:47082", + "right": "pull_request:47210", "accept": false, - "reason": "MTP docs and linear-attention dispatch simplification are unrelated generation/backend changes." + "reason": "RoMa keypoint matcher support versus packed SDPA varlen routing; unrelated." }, { "left": "pull_request:46196", "right": "pull_request:47461", "accept": false, - "reason": "SDPA dispatch support and stale phonemizer language fixes are unrelated." + "reason": "Audio SDPA dispatch versus stale phonemizer language fix; different subsystems." }, { "left": "pull_request:47204", "right": "pull_request:47422", "accept": false, - "reason": "Exporter test image/dependency setup and torchaudio pinning/RNNT cleanup are different CI fixes." - }, - { - "left": "pull_request:47113", - "right": "pull_request:47452", - "accept": false, - "reason": "Audio feature requirement fixes and mamba/causal-conv standardization are unrelated." - }, - { - "left": "pull_request:47336", - "right": "pull_request:47357", - "accept": false, - "reason": "Model saving offload bugfix and FSDP CI/save work overlap only on 'saving'; they target different failure modes." + "reason": "Exporter/CI image dependency changes versus quantization docker torch pinning; related only at the CI level, not the same bug." }, { - "left": "pull_request:46196", - "right": "pull_request:47450", + "left": "pull_request:47108", + "right": "pull_request:47204", "accept": false, - "reason": "SDPA dispatch support and linear-attention function dispatch simplification are different attention subsystems." + "reason": "Doc-builder comment update versus docker exporter-test dependency additions; unrelated." } ] }, @@ -25213,34 +25261,33 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-96", + "cluster_id": "cluster-40546-97", "nodes": [ "pull_request:41855", "pull_request:43424", "pull_request:43451", - "pull_request:43823", "pull_request:44517", "pull_request:45157", "pull_request:45273", "pull_request:46196", + "pull_request:46266", "pull_request:46381", "pull_request:46556", "pull_request:46604", "pull_request:46658", + "pull_request:46836", "pull_request:46912", "pull_request:46957", "pull_request:47077", "pull_request:47078", "pull_request:47096", "pull_request:47105", - "pull_request:47108", "pull_request:47113", "pull_request:47143", "pull_request:47163", "pull_request:47166", "pull_request:47175", "pull_request:47179", - "pull_request:47204", "pull_request:47210", "pull_request:47216", "pull_request:47243", @@ -25262,23 +25309,23 @@ "pull_request:47451", "pull_request:47452", "pull_request:47457", - "pull_request:47461" + "pull_request:47461", + "pull_request:47462" ], "soft_pairs": [ "pull_request:47360|pull_request:47422", - "pull_request:46912|pull_request:47078", - "pull_request:47108|pull_request:47204", "pull_request:47336|pull_request:47447", "pull_request:47179|pull_request:47301", "pull_request:47175|pull_request:47461", "pull_request:47322|pull_request:47451", "pull_request:47291|pull_request:47451", + "pull_request:46556|pull_request:46836", "pull_request:47216|pull_request:47336", "pull_request:47166|pull_request:47422", - "pull_request:47298|pull_request:47438", "pull_request:47275|pull_request:47422", - "pull_request:47143|pull_request:47320", + "pull_request:47298|pull_request:47438", "pull_request:46912|pull_request:47294", + "pull_request:47143|pull_request:47320", "pull_request:47175|pull_request:47294", "pull_request:47175|pull_request:47438", "pull_request:47216|pull_request:47243", @@ -25317,6 +25364,7 @@ "pull_request:47113|pull_request:47275", "pull_request:47077|pull_request:47367", "pull_request:45273|pull_request:47447", + "pull_request:46266|pull_request:47452", "pull_request:45273|pull_request:47294", "pull_request:47294|pull_request:47306", "pull_request:47210|pull_request:47367", @@ -25326,349 +25374,337 @@ "pull_request:46381|pull_request:47096", "pull_request:47166|pull_request:47457", "pull_request:45273|pull_request:47243", - "pull_request:41855|pull_request:43424", - "pull_request:41855|pull_request:43823" + "pull_request:47322|pull_request:47462", + "pull_request:41855|pull_request:43424" ], - "prepared_review_unit_hash": "c6104929d516986cd551d4bdd47f1ea78e4cb38322daa891de4c6418bead090e", + "prepared_review_unit_hash": "a011fd171f8e6d0945ffaa65bd93cd3cc82878cdc238f87b436266aecfb8b608", "component_budget": { - "node_count": 96, - "item_count": 96, - "soft_pair_count": 272, - "serialized_chars": 334252, - "estimated_input_tokens": 83563, - "estimated_eval_tokens": 167382 + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 63, - "serialized_chars": 134477, - "estimated_input_tokens": 33620, - "estimated_eval_tokens": 67496 + "serialized_chars": 137099, + "estimated_input_tokens": 34275, + "estimated_eval_tokens": 68806 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is not a duplicate cluster: the PRs span unrelated model additions, docs translations, CI/test tweaks, export/quantization work, and separate low-level runtime fixes. None of the soft edges look like the same underlying bug or change, so all are rejected.", - "confidence": 0.98, + "summary": "The cluster is mostly a set of unrelated PRs with superficial similarity from shared subsystems like model additions, export, generation, docs, and CI. None of the soft-edge pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: the bucket mixes unrelated model adds, docs-only edits, CI updates, exporter/runtime fixes, and test-only changes, so any representative would be arbitrary.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "No global best PR for deduping; the more substantive PRs still target distinct code paths and cannot plausibly be merged into one change.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:47360", "right": "pull_request:47422", "accept": false, - "reason": "CI/OTEL workflow change vs Docker pinning for exporter tests; different subsystems and different bugs." - }, - { - "left": "pull_request:46912", - "right": "pull_request:47078", - "accept": false, - "reason": "RoPE test coverage vs Japanese documentation wording; unrelated changes." - }, - { - "left": "pull_request:47108", - "right": "pull_request:47204", - "accept": false, - "reason": "Doc-builder comment tweak vs adding executorch/onnx to torch-light; not the same fix." + "reason": "Different CI concerns: OTEL workflow instrumentation vs torch/torchaudio pinning and quantization image fixes." }, { "left": "pull_request:47336", "right": "pull_request:47447", "accept": false, - "reason": "Fully offloaded save bug vs sliding-window rollback for speculative decoding; separate cache/save paths." + "reason": "One fixes fully offloaded model saving; the other adds sliding-window cache rollback for speculative decoding. Different code paths." }, { "left": "pull_request:47179", "right": "pull_request:47301", "accept": false, - "reason": "Granite model support vs assisted-decoding docs for MTP; unrelated feature areas." + "reason": "New Granite-SWA model support is unrelated to assisted decoding docs for multi-token prediction." }, { "left": "pull_request:47175", "right": "pull_request:47461", "accept": false, - "reason": "Olmo3 CI expectation updates vs Wav2Vec2 phonemizer state bug; unrelated." + "reason": "CI expectation updates for Olmo3 are unrelated to the stale phonemizer language bug in Wav2Vec2Phoneme." }, { "left": "pull_request:47322", "right": "pull_request:47451", "accept": false, - "reason": "Quark config-class compatibility fix vs attention value padding; different code paths." + "reason": "Quark config/version handling is unrelated to moving value padding into attention interfaces." }, { "left": "pull_request:47291", "right": "pull_request:47451", "accept": false, - "reason": "Generation config deprecation vs attention interface padding; not the same issue." + "reason": "Deprecating ContinuousBatchingConfig from generation config is a different change from attention value padding." + }, + { + "left": "pull_request:46556", + "right": "pull_request:46836", + "accept": false, + "reason": "Image size utility support for PIL images is not the same change as adding Unlimited OCR model support." }, { "left": "pull_request:47216", "right": "pull_request:47336", "accept": false, - "reason": "Compressed-tensors quantization support vs offloaded model saving; different functionality." + "reason": "Compressed-tensors FP8 integration is unrelated to offloaded model saving behavior." }, { "left": "pull_request:47166", "right": "pull_request:47422", "accept": false, - "reason": "Quantization test extensions vs Docker pinning for exporter CI; unrelated." - }, - { - "left": "pull_request:47298", - "right": "pull_request:47438", - "accept": false, - "reason": "MI450 model failures / image decoder fallback vs mask-materialization reduction; distinct runtime problems." + "reason": "Quantization test tweaks for GPTQ/MXFP4 are unrelated to the CI torch pinning fix." }, { "left": "pull_request:47275", "right": "pull_request:47422", "accept": false, - "reason": "EfficientViT-SAM model addition vs CI image pinning; no shared underlying bug." + "reason": "Adding EfficientVit-SAM is a new model PR, not the same as fixing CI torch/torchaudio versioning." }, { - "left": "pull_request:47143", - "right": "pull_request:47320", + "left": "pull_request:47298", + "right": "pull_request:47438", "accept": false, - "reason": "NeuCodec model support vs response_schema deprecation; unrelated API/model work." + "reason": "MI450 model failures and less-materialized masks both touch runtime behavior, but the concrete problems and fixes are different." }, { "left": "pull_request:46912", "right": "pull_request:47294", "accept": false, - "reason": "RoPE tests vs kernel-native causal_conv1d functions; different code paths and purposes." + "reason": "RoPE test coverage and kernel-native causal_conv1d functions are separate infrastructure changes." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47320", + "accept": false, + "reason": "NeuCodec model support is unrelated to deprecating the old response schema in CLI serving." }, { "left": "pull_request:47175", "right": "pull_request:47294", "accept": false, - "reason": "Olmo3 CI updates vs causal_conv1d kernels; unrelated." + "reason": "Olmo3 CI output updates do not match the causal_conv1d kernel integration work." }, { "left": "pull_request:47175", "right": "pull_request:47438", "accept": false, - "reason": "Olmo3 CI updates vs less-materialized masks; different bugs and scopes." + "reason": "CI expectation updates for a model test are not the same as boolean mask materialization changes." }, { "left": "pull_request:47216", "right": "pull_request:47243", "accept": false, - "reason": "Compressed-tensors fp8 vs Executorch exporter fixes; both exporter-related, but separate concrete problems." + "reason": "Compressed-tensors FP8 work and Executorch exporter fixes are both infra-related but fix different mechanisms." }, { "left": "pull_request:47077", "right": "pull_request:47461", "accept": false, - "reason": "Japanese docs translation vs tokenizer language-state bug; unrelated." + "reason": "Japanese docs translation edits are unrelated to the Wav2Vec2Phoneme language-state bug." }, { "left": "pull_request:45273", "right": "pull_request:47451", "accept": false, - "reason": "Liger eval OOM fix vs attention padding logic; not the same change." + "reason": "Liger eval OOM fix in Trainer is unrelated to attention interface value padding." }, { "left": "pull_request:47216", "right": "pull_request:47294", "accept": false, - "reason": "Quantization format support vs causal_conv1d kernel utilities; unrelated." + "reason": "Compressed-tensors integration is unrelated to kernel-native causal_conv1d support." }, { "left": "pull_request:47078", "right": "pull_request:47452", "accept": false, - "reason": "Japanese wording/docs cleanup vs model implementation alignment work; unrelated." + "reason": "Japanese wording/doc updates are unrelated to the broader Mamba alignment refactor." }, { "left": "pull_request:47175", "right": "pull_request:47447", "accept": false, - "reason": "Olmo3 CI expectations vs speculative-decoding cache rollback; different concerns." + "reason": "Model CI expectation churn and speculative-decoding cache rollback are different issues." }, { "left": "pull_request:46912", "right": "pull_request:47438", "accept": false, - "reason": "RoPE test coverage vs recurrent-mask materialization reduction; both low-level, but not the same bug." + "reason": "RoPE test coverage and less-materialized masks are different attention-related changes, not one bug." }, { "left": "pull_request:47105", "right": "pull_request:47298", "accept": false, - "reason": "Re-enable VITS test vs MI450 failures; test coverage change vs runtime bugfix." + "reason": "Re-enabling a VITS test is unrelated to MI450 hardware model failures." }, { "left": "pull_request:47078", "right": "pull_request:47322", "accept": false, - "reason": "Japanese docs wording vs Quark config changes; unrelated." + "reason": "Japanese doc wording is unrelated to Quark config class changes." }, { "left": "pull_request:47322", "right": "pull_request:47461", "accept": false, - "reason": "Quark version/config changes vs stale phonemizer language handling; unrelated." + "reason": "Quark quantization config changes are unrelated to the stale phonemizer language fix." }, { "left": "pull_request:45157", "right": "pull_request:46658", "accept": false, - "reason": "PrismML support vs Step-3.7 mini-model tooling; different model families and goals." + "reason": "Both are large model-support efforts, but they target different architectures and codepaths with no shared bug/change." }, { "left": "pull_request:44517", "right": "pull_request:47301", "accept": false, - "reason": "Qwen3 TTS model support vs MTP documentation; unrelated." + "reason": "Qwen3-TTS model support is unrelated to assisted-decoding docs for MTP." }, { "left": "pull_request:47294", "right": "pull_request:47438", "accept": false, - "reason": "Causal-conv kernel plumbing vs mask materialization optimization; separate infrastructure changes." + "reason": "Kernel-native causal_conv1d support and less-materialized masks are distinct optimization paths." }, { "left": "pull_request:47113", - "right": "pull_request:47304", - "accept": false, - "reason": "Audio feature/processor requirement fixes vs docs-only MPS graph cache cleanup; unrelated." - }, - { - "left": "pull_request:46912", - "right": "pull_request:47291", - "accept": false, - "reason": "RoPE tests vs generation-config deprecation; unrelated." - }, - { - "left": "pull_request:47438", - "right": "pull_request:47447", - "accept": false, - "reason": "Less-materialized masks vs sliding-window rollback for speculative decoding; both generation internals, but different concrete mechanisms." - }, - { - "left": "pull_request:43451", - "right": "pull_request:44517", - "accept": false, - "reason": "Molmo2 model addition vs Qwen3 TTS model addition; different model families." - }, - { - "left": "pull_request:45273", - "right": "pull_request:46196", - "accept": false, - "reason": "Liger eval OOM fix vs SDPA dispatch support; separate runtime optimizations." - }, - { - "left": "pull_request:46556", - "right": "pull_request:47301", + "right": "pull_request:47243", "accept": false, - "reason": "Processor updates across models vs MTP docs; unrelated." + "reason": "Audio feature/proc requirement decorators are unrelated to Executorch exporter fixes." }, { "left": "pull_request:44517", "right": "pull_request:47143", "accept": false, - "reason": "Qwen3 TTS vs NeuCodec; distinct model integrations." + "reason": "Qwen3-TTS support and NeuCodec support are different model additions." }, { "left": "pull_request:44517", "right": "pull_request:47163", "accept": false, - "reason": "Qwen3 TTS support vs XPU paged-attention tests; unrelated." + "reason": "Qwen3-TTS model support is unrelated to XPU paged-attention test coverage." }, { "left": "pull_request:47210", "right": "pull_request:47291", "accept": false, - "reason": "Packed SDPA varlen routing vs generation-config deprecation; unrelated." + "reason": "Packed SDPA varlen routing and generation config deprecation are unrelated." }, { "left": "pull_request:46604", "right": "pull_request:47163", "accept": false, - "reason": "Mistral tekken tokenizer utilities vs XPU paged-attention tests; unrelated." + "reason": "Mistral tekken tokenizer utilities are unrelated to XPU paged-attention tests." }, { "left": "pull_request:47113", "right": "pull_request:47175", "accept": false, - "reason": "Audio backend requirement fixes vs Olmo3 CI expectations; unrelated." + "reason": "Audio requirement fixes do not match Olmo3 CI expectation updates." }, { "left": "pull_request:43451", "right": "pull_request:47451", "accept": false, - "reason": "Molmo2 model support vs flash-attention value padding; different code paths." + "reason": "Molmo2 model support and attention value padding are completely different changes." }, { "left": "pull_request:43451", "right": "pull_request:46556", "accept": false, - "reason": "Molmo2 model support vs processors update sweep; unrelated." + "reason": "Molmo2 model support is unrelated to PIL image size handling in image_utils." }, { "left": "pull_request:46556", "right": "pull_request:47096", "accept": false, - "reason": "Processor sweep vs Kimi exportability work; different areas." + "reason": "Image utility changes are unrelated to Exportable Kimi exporter support." }, { "left": "pull_request:46957", "right": "pull_request:47360", "accept": false, - "reason": "Finegrained FP8 kernel/generation changes vs CI OTEL workflow; unrelated." + "reason": "Finegrained FP8 fused-kernel logic is unrelated to the OTEL CI workflow change." }, { "left": "pull_request:45273", "right": "pull_request:47447", "accept": false, - "reason": "Liger eval OOM fix vs cache rollback for speculative decoding; separate generation behaviors." + "reason": "Trainer eval OOM fix and cache rollback for speculative decoding are different generation/training paths." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47452", + "accept": false, + "reason": "DINO modernization work and Mamba coverage refactors are separate model-family changes." }, { "left": "pull_request:45273", "right": "pull_request:47294", "accept": false, - "reason": "Liger eval OOM fix vs causal_conv1d kernel support; unrelated." + "reason": "Trainer eval logits handling and causal_conv1d kernel integration are unrelated." }, { "left": "pull_request:47294", "right": "pull_request:47306", "accept": false, - "reason": "Kernel-native causal_conv1d functions vs Neuron experiments script; unrelated." + "reason": "Kernel-native causal_conv1d functions and Neuron experiment scripts are unrelated." }, { - "left": "pull_request:47210", + "left": "pull_request:47077", "right": "pull_request:47367", "accept": false, - "reason": "Packed SDPA varlen attention vs pipeline device-placement docs/code; different subsystems." + "reason": "Japanese translation edits are unrelated to pipeline device-placement clarification." }, { - "left": "pull_request:47275", - "right": "pull_request:47457", + "left": "pull_request:46957", + "right": "pull_request:47336", + "accept": false, + "reason": "Finegrained FP8 decoding optimizations and fully offloaded model saving are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46381", "accept": false, - "reason": "EfficientViT-SAM model support vs Inkling test autofix; unrelated." + "reason": "Molmo2 model support is unrelated to allowing tokenizers 0.23.1." + }, + { + "left": "pull_request:46381", + "right": "pull_request:47096", + "accept": false, + "reason": "Tokenizer dependency/version compatibility changes are unrelated to Exportable Kimi." }, { "left": "pull_request:47166", "right": "pull_request:47457", "accept": false, - "reason": "Quantization test changes vs Inkling test autofix; unrelated test-only work." + "reason": "Quantization test extension and Inkling test fixes are different test suites for different subsystems." }, { - "left": "pull_request:41855", - "right": "pull_request:43424", + "left": "pull_request:45273", + "right": "pull_request:47243", + "accept": false, + "reason": "Liger eval OOM behavior and Executorch exporter fixes are not the same code-path problem." + }, + { + "left": "pull_request:47322", + "right": "pull_request:47462", "accept": false, - "reason": "Mistral tokenizer method addition vs ExecuTorch dynamic-shape test; unrelated." + "reason": "Quark config changes and MTP support for GLM OCR/MROPEs are unrelated features." }, { "left": "pull_request:41855", - "right": "pull_request:43823", + "right": "pull_request:43424", "accept": false, - "reason": "Mistral tokenizer fix vs MobileLLM model support; different features and code paths." + "reason": "Mistral missing tokenizer methods and Executorch dynamic-shape tests do not fix the same problem." } ] }, @@ -25690,8 +25726,9 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-96", + "cluster_id": "cluster-40546-97", "nodes": [ + "pull_request:41855", "pull_request:42668", "pull_request:43251", "pull_request:43424", @@ -25732,7 +25769,6 @@ "pull_request:47310", "pull_request:47322", "pull_request:47336", - "pull_request:47367", "pull_request:47380", "pull_request:47387", "pull_request:47443", @@ -25742,6 +25778,7 @@ "pull_request:47461" ], "soft_pairs": [ + "pull_request:41855|pull_request:43823", "pull_request:42668|pull_request:43823", "pull_request:43424|pull_request:47301", "pull_request:43823|pull_request:45655", @@ -25796,353 +25833,261 @@ "pull_request:46509|pull_request:47143", "pull_request:46509|pull_request:47163", "pull_request:45157|pull_request:46836", - "pull_request:46836|pull_request:47387", "pull_request:47291|pull_request:47322", + "pull_request:46836|pull_request:47387", "pull_request:45157|pull_request:46509", - "pull_request:47301|pull_request:47336", - "pull_request:47304|pull_request:47367", - "pull_request:46509|pull_request:47387" + "pull_request:47301|pull_request:47336" ], - "prepared_review_unit_hash": "37cfdf546ea1d607f9641679f984ce15976ae82cd0822e22b32243f39d96b9a9", + "prepared_review_unit_hash": "c6f570b00e017e0f1ae191eebff1baf90d3d8c998d5141191d682b9b29e5ef88", "component_budget": { - "node_count": 96, - "item_count": 96, - "soft_pair_count": 272, - "serialized_chars": 334252, - "estimated_input_tokens": 83563, - "estimated_eval_tokens": 167382 + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 60, - "serialized_chars": 128370, - "estimated_input_tokens": 32093, - "estimated_eval_tokens": 64442 + "soft_pair_count": 59, + "serialized_chars": 128019, + "estimated_input_tokens": 32005, + "estimated_eval_tokens": 64266 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many pairs are only superficially similar (same subsystem or both test/doc updates), while a few are true duplicates around the same exact bug/fix pattern. The clearest duplicate groups are the Wav2Vec2 phonemizer state fix, the MiniMax-M3-VL OOM/load_error test fixes, the Mamba2 OOM test fixes, and one DeepSeek model-file cleanup pair.", - "confidence": 0.66, + "summary": "This is a loose cluster of mostly unrelated PRs. The only clear duplicate groups are the Wav2Vec2Phoneme tokenizer stale-language fix, the MiniMax-M3-VL OOM/offload test fix, the Mamba2 OOM test fix, and the DeepSeek v3/v32 code-cleanup trio. Most other pairs are different features, models, or docs work and should stay separate.", + "confidence": 0.64, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46687 is the cleanest duplicate anchor: it is a surgical one-file fix, has an explicit issue target (46614), and matches its twin PR almost exactly.", + "canonical_pr_reason": "47443 is the best canonical PR anchor because it contains the core shared DeepSeek v3/v32 model-code change that the other true duplicate PRs mirror, while the remaining accepted groups are smaller follow-up test fixes.", "best_issue_reason": null, - "best_pr_reason": "PR 46687 is the best representative of the duplicate-worthy subset because it is precise, issue-linked, and clearly tied to a single concrete bug in the tokenizer state handling.", + "best_pr_reason": "47443 is the strongest overall representative of the genuine duplicate subcluster: it carries the shared code-path fix across multiple model files and is more central than the smaller test-only variants.", "soft_edge_verdicts": [ + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Unrelated topics: Mistral tokenizer API gap vs a new MobileLLM model addition." + }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "Different work: processor auto-loading robustness vs adding a new model implementation." + "reason": "Processor robustness fix and MobileLLM model addition are different changes." }, { "left": "pull_request:43424", "right": "pull_request:47301", "accept": false, - "reason": "Unrelated fixes: ExecuTorch test coverage vs assisted-decoding docs for MTP." + "reason": "ExecuTorch dynamic-shape test coverage and MTP docs are unrelated." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New model support vs config-resolution bugfix; not the same change." + "reason": "New model support vs config-resolution ordering fix; different code paths." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 test fixes vs a LayerNorm spelling/renaming cleanup; different bugs." + "reason": "Mistral4 test fixes and CLIP-like layernorm rename are unrelated." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "Different models and failures; only broad test-fix similarity." + "reason": "ExecuTorch export test and Mistral4 test fixes are different model/runtime issues." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "Model addition vs Falcon refactor; no shared bug path." + "reason": "Model addition and Falcon output-interface refactor are not the same bug/change." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Cross-entropy kwargs fix vs new model support; unrelated." + "reason": "cross_entropy kwargs passthrough and MobileLLM support are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 test noise vs TDT loss kernel support; not the same concrete issue." + "reason": "Mistral4 test work and TDT loss kernel support do not share a concrete code path." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "ExecuTorch test coverage vs MobileLLM model addition; unrelated." + "reason": "ExecuTorch export tests and MobileLLM model addition are unrelated." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Loss utility kwargs fix vs Mistral4 test updates; different code paths." + "reason": "Loss-function kwargs fix is unrelated to Mistral4 test failures." }, { "left": "pull_request:44827", "right": "pull_request:45655", "accept": false, - "reason": "Test/model issue vs config-resolution bugfix; not duplicate content." + "reason": "These touch different subsystems and different failure modes." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon output-interface refactor vs model-creation skill; unrelated." + "reason": "Falcon refactor and model-creation skill docs are not the same change." }, { "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 tests vs repo skill scaffolding; not the same bug/change." + "reason": "Model tests and a repo skill scaffold are unrelated." }, { - "left": "pull_request:43995", - "right": "pull_request:44827", + "left": "pull_request:43823", + "right": "pull_request:45213", "accept": false, - "reason": "Different model families and different objectives; no concrete overlap." + "reason": "MobileLLM model support and skill/docs scaffolding are different work." }, { - "left": "pull_request:43823", - "right": "pull_request:45213", + "left": "pull_request:45157", + "right": "pull_request:46836", "accept": false, - "reason": "New model support vs skill/documentation scaffolding; unrelated." + "reason": "PrismML Bonsai support and Unlimited OCR are distinct model additions." }, { "left": "pull_request:46687", "right": "pull_request:47461", "accept": true, - "reason": "Same tokenizer file and same stale-phonemizer-lang bug; clearly the same fix with a wording/packaging tweak." + "reason": "Same tokenizer bug in the same file; both fix stale phonemizer_lang state and can be merged together." }, { "left": "pull_request:47123", "right": "pull_request:47280", "accept": true, - "reason": "Both address the same MiniMax-M3-VL load_error/OOM test instability by adding offload-folder handling." + "reason": "Same MiniMax-M3-VL load_error/offload-folder test fix in the same test file." }, { "left": "pull_request:47266", "right": "pull_request:47443", "accept": false, - "reason": "Different model failures and different fixes; shared style is not enough to merge them." + "reason": "Different model families and different fix styles; not the same concrete bug/change." }, { "left": "pull_request:47282", "right": "pull_request:47380", "accept": true, - "reason": "Same Mamba2 OOM test issue; both patches are test-level mitigations for the same failure mode." + "reason": "Same Mamba2 OOM test issue in the same integration test, with only minor wording/setup differences." }, { "left": "pull_request:47123", "right": "pull_request:47443", "accept": false, - "reason": "Different models and failure modes; one is MiniMax-M3-VL, the other is DeepSeek_v32 cleanup." + "reason": "MiniMax-M3-VL load_error and DeepSeek/MPT refactoring are unrelated." }, { "left": "pull_request:47123", "right": "pull_request:47266", "accept": false, - "reason": "MiniMax-M3-VL OOM fix vs DeepSeek_v32 load_error fix; not the same code-path problem." + "reason": "Different models and different failing tests; not mergeable as one fix." }, { "left": "pull_request:47266", "right": "pull_request:47280", "accept": false, - "reason": "DeepSeek_v32 and MiniMax-M3-VL are unrelated model failures; similar test plumbing only." - }, - { - "left": "pull_request:47124", - "right": "pull_request:47310", - "accept": false, - "reason": "Falcon-Mamba expected-output updates are unrelated to the Gemma3 cache-implementation fix." + "reason": "DeepSeek V3.2 load_error and MiniMax-M3-VL load_error are separate model issues." }, { "left": "pull_request:47280", - "right": "pull_request:47310", + "right": "pull_request:47443", "accept": false, - "reason": "Different model families and different bug classes." + "reason": "MiniMax-M3-VL test fix is unrelated to the DeepSeek code cleanup." }, { - "left": "pull_request:47380", + "left": "pull_request:47443", "right": "pull_request:47444", - "accept": false, - "reason": "Mamba2 OOM test fix vs MPT load_error fix; unrelated." - }, - { - "left": "pull_request:47122", - "right": "pull_request:47281", - "accept": false, - "reason": "EdgeTam import-or-config fix vs GLM46V model refactor; not the same issue." + "accept": true, + "reason": "They share the same DeepSeek model-file cleanup and look like parallel variants of the same code-path fix." }, { - "left": "pull_request:47266", + "left": "pull_request:47443", "right": "pull_request:47445", - "accept": false, - "reason": "DeepSeek_v32 test offload handling vs PEFT hotswap/load logic; different changes." - }, - { - "left": "pull_request:47280", - "right": "pull_request:47444", - "accept": false, - "reason": "MiniMax-M3-VL load fix vs MPT integration fix; not duplicate content." - }, - { - "left": "pull_request:47123", - "right": "pull_request:47444", - "accept": false, - "reason": "Different model failures and different remedies." - }, - { - "left": "pull_request:47281", - "right": "pull_request:47443", - "accept": false, - "reason": "GLM46V refactor is unrelated to DeepSeek_v32 load-error cleanup." - }, - { - "left": "pull_request:47122", - "right": "pull_request:47310", - "accept": false, - "reason": "EdgeTam backbone kwargs cleanup vs Gemma3 generation-config behavior; unrelated." - }, - { - "left": "pull_request:47173", - "right": "pull_request:47449", - "accept": false, - "reason": "XPU expectation updates for vision models vs Qwen3 Omni MoE model fix; different bugs." - }, - { - "left": "pull_request:47292", - "right": "pull_request:47449", - "accept": false, - "reason": "Different models and different test-failure causes." + "accept": true, + "reason": "Same shared DeepSeek model-file change; the PEFT-related wrapper is a follow-up around the same fix." }, { - "left": "pull_request:47173", - "right": "pull_request:47292", - "accept": false, - "reason": "Both are XPU-related test updates, but for different models and different failure modes." + "left": "pull_request:47444", + "right": "pull_request:47445", + "accept": true, + "reason": "Both carry the same DeepSeek model-code cleanup and could plausibly be combined." }, { - "left": "pull_request:46912", - "right": "pull_request:47175", + "left": "pull_request:47291", + "right": "pull_request:47322", "accept": false, - "reason": "RoPE behavior refactor vs an OLMo3 CI expectation update; not the same concrete fix." + "reason": "Deprecating continuous batching config and Quark config-class changes are separate quantization issues." }, { "left": "pull_request:47301", "right": "pull_request:47304", "accept": false, - "reason": "Assisted-decoding docs vs MPS graph-cache docs; unrelated." - }, - { - "left": "pull_request:47216", - "right": "pull_request:47301", - "accept": false, - "reason": "Compressed-tensors fp8 support vs MTP docs; different topics." - }, - { - "left": "pull_request:47077", - "right": "pull_request:47304", - "accept": false, - "reason": "Japanese DistilBERT docs vs MPS graph-cache docs; unrelated." - }, - { - "left": "pull_request:46658", - "right": "pull_request:46836", - "accept": false, - "reason": "Two separate new-model efforts (Step3p7 vs Unlimited OCR); not the same model or bug." - }, - { - "left": "pull_request:47210", - "right": "pull_request:47301", - "accept": false, - "reason": "Packed SDPA routing change vs MTP docs; not a duplicate." - }, - { - "left": "pull_request:47163", - "right": "pull_request:47166", - "accept": false, - "reason": "Paged-attention XPU test support vs GPTQ test extensions; different test areas and failure modes." + "reason": "MTP generation docs and MPS graph-cache docs are unrelated documentation work." }, { "left": "pull_request:47003", - "right": "pull_request:47243", - "accept": false, - "reason": "OpenVINO exporter introduction vs ExecuTorch exporter fixes; same broad area, different concrete changes." - }, - { - "left": "pull_request:46509", - "right": "pull_request:47143", - "accept": false, - "reason": "Nemotron Omni support vs NeuCodec support; different model families." - }, - { - "left": "pull_request:45157", - "right": "pull_request:46836", - "accept": false, - "reason": "PrismML Bonsai support vs Unlimited OCR support; unrelated model additions." - }, - { - "left": "pull_request:45157", - "right": "pull_request:46509", + "right": "pull_request:47243", "accept": false, - "reason": "Different model-support PRs with no shared code-path." + "reason": "OpenVINO exporter work and ExecuTorch exporter fixes are different exporter implementations." }, { - "left": "pull_request:47291", - "right": "pull_request:47322", + "left": "pull_request:47216", + "right": "pull_request:47301", "accept": false, - "reason": "Generation-config deprecation vs AMD Quark config-class changes; not the same bug." + "reason": "Compressed-tensors fp8 and MTP docs are unrelated changes." }, { - "left": "pull_request:47301", - "right": "pull_request:47336", + "left": "pull_request:47077", + "right": "pull_request:47304", "accept": false, - "reason": "MTP docs vs fully-offloaded model saving bugfix; unrelated." + "reason": "Japanese translation edits and MPS graph-cache docs are not the same artifact." }, { - "left": "pull_request:47304", - "right": "pull_request:47367", + "left": "pull_request:46509", + "right": "pull_request:47143", "accept": false, - "reason": "MPS graph-cache docs vs pipeline device-placement behavior; not the same change." + "reason": "Nemotron Omni support and NeuCodec support are separate model additions." }, { - "left": "pull_request:46509", + "left": "pull_request:46836", "right": "pull_request:47387", "accept": false, - "reason": "Nemotron Omni support vs one-to-many conversion support; different features." + "reason": "Unlimited OCR support and one-to-many conversion loading are unrelated." }, { - "left": "pull_request:47443", - "right": "pull_request:47444", - "accept": true, - "reason": "Same shared DeepSeek-family model-file cleanup across the same code paths; the only difference is the test target, so this could plausibly be one PR." + "left": "pull_request:47163", + "right": "pull_request:47166", + "accept": false, + "reason": "Paged-attention XPU tests and quantization test extensions are different issues." }, { - "left": "pull_request:47444", - "right": "pull_request:47445", + "left": "pull_request:47210", + "right": "pull_request:47301", "accept": false, - "reason": "Overlap in touched model files is not enough; one is a PEFT hotswap fix and the other is an integration-test/load-error fix." + "reason": "Packed SDPA routing and MTP docs are unrelated." }, { - "left": "pull_request:47443", - "right": "pull_request:47445", + "left": "pull_request:47003", + "right": "pull_request:47336", "accept": false, - "reason": "Same broad model family, but distinct fixes: model-file cleanup vs PEFT load behavior." + "reason": "Exporter feature work and offloading save bugfix are different code paths." } ] }, @@ -26164,7 +26109,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-40546-96", + "cluster_id": "cluster-40546-97", "nodes": [ "pull_request:40546", "pull_request:43451", @@ -26183,6 +26128,7 @@ "pull_request:47294", "pull_request:47298", "pull_request:47301", + "pull_request:47304", "pull_request:47306", "pull_request:47322", "pull_request:47367", @@ -26190,6 +26136,8 @@ "pull_request:47452" ], "soft_pairs": [ + "pull_request:47304|pull_request:47367", + "pull_request:46509|pull_request:47387", "pull_request:46912|pull_request:47298", "pull_request:46509|pull_request:47301", "pull_request:47175|pull_request:47298", @@ -26216,7 +26164,6 @@ "pull_request:43451|pull_request:47003", "pull_request:47003|pull_request:47096", "pull_request:43451|pull_request:47322", - "pull_request:47096|pull_request:47322", "pull_request:43451|pull_request:47306", "pull_request:47096|pull_request:47306", "pull_request:47003|pull_request:47306", @@ -26225,238 +26172,244 @@ "pull_request:47096|pull_request:47143", "pull_request:47096|pull_request:47163" ], - "prepared_review_unit_hash": "aab9575bb887e9598f3004f5a237789a439fa05dc95c59664c3b6c91957157d7", + "prepared_review_unit_hash": "0c0d47f19b70de9deb41ab386142f993ac7ca9ecd0a271f2b1282bd7848129f2", "component_budget": { - "node_count": 96, - "item_count": 96, - "soft_pair_count": 272, - "serialized_chars": 334252, - "estimated_input_tokens": 83563, - "estimated_eval_tokens": 167382 + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 }, "budget": { - "node_count": 22, - "item_count": 22, - "soft_pair_count": 34, - "serialized_chars": 62318, - "estimated_input_tokens": 15580, - "estimated_eval_tokens": 31416 + "node_count": 23, + "item_count": 23, + "soft_pair_count": 35, + "serialized_chars": 64456, + "estimated_input_tokens": 16114, + "estimated_eval_tokens": 32484 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a heterogeneous mix of unrelated model additions, exporter/integration work, and isolated test/doc fixes. None of the proposed soft edges look like true duplicates of the same underlying change.", - "confidence": 0.97, + "summary": "This cluster is a loose semantic grouping of unrelated PRs: new model integrations, exporter/runtime work, docs-only changes, and test/CI fixes. None of the soft edges look like the same concrete bug or change, so they should all remain separate.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No clear canonical PR: the items span independent features and fixes across different models and subsystems, so there is no single representative duplicate target.", + "canonical_pr_reason": "There is no true canonical PR for this cluster because the items are unrelated; if a representative has to be named, 43451 ('Add Molmo2') is the most complete self-contained feature PR (docs + tests + model plumbing), but it does not subsume the others.", "best_issue_reason": null, - "best_pr_reason": "No PR is globally suitable as a canonical duplicate because the cluster is not one change; it is a grab bag of unrelated PRs.", + "best_pr_reason": "43451 is the best representative anchor only in the weak sense that it is a substantial, self-contained implementation PR. It is not a duplicate of the other items, but it is the clearest standalone feature PR in the set.", "soft_edge_verdicts": [ + { + "left": "pull_request:47304", + "right": "pull_request:47367", + "accept": false, + "reason": "Both are docs-related, but one is about MPS graph cache docs and the other changes pipeline device placement semantics; different changes." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47387", + "accept": false, + "reason": "Nemotron Omni model support vs one-to-many conversion infrastructure; different layers and different code paths." + }, { "left": "pull_request:46912", "right": "pull_request:47298", "accept": false, - "reason": "Different test updates for different failures/model suites; not the same bug." + "reason": "RoPE test expansion across models vs MI450 failure fixes; only a broad testing theme, not the same change." }, { "left": "pull_request:46509", "right": "pull_request:47301", "accept": false, - "reason": "Nemotron Omni model support vs docs-only MTP support; unrelated changes." + "reason": "Nemotron Omni support and MTP docs are unrelated feature areas." }, { "left": "pull_request:47175", "right": "pull_request:47298", "accept": false, - "reason": "Separate XPU/CUDA expectation edits in different model tests; not one underlying fix." + "reason": "Olmo3 CI expectation updates vs MI450 model/runtime fixes; not the same underlying bug or change." }, { "left": "pull_request:47163", "right": "pull_request:47210", "accept": false, - "reason": "Paged-attention XPU test coverage vs packed-SDPA varlen attention implementation; different code paths." + "reason": "Both concern attention/XPU, but 47163 is test coverage and 47210 is runtime routing to varlen attention; different concrete changes." }, { "left": "pull_request:45157", "right": "pull_request:47387", "accept": false, - "reason": "Prism GGUF dequantization support vs one-to-many conversion plumbing; unrelated." + "reason": "Prism GGUF dequantization support vs generic one-to-many conversion support; unrelated." }, { "left": "pull_request:47294", "right": "pull_request:47298", "accept": false, - "reason": "Kernel native causal_conv1d work vs MI450 model-failure fixes; different problems." + "reason": "Native causal_conv1d kernel plumbing vs MI450 model failures; different problems." }, { "left": "pull_request:46658", "right": "pull_request:47387", "accept": false, - "reason": "Step3.7 Flash model support vs conversion framework changes; not the same change." + "reason": "Step3.7 checkpoint conversion tooling vs core one-to-many conversion support; related only at a very high level." }, { "left": "pull_request:47143", "right": "pull_request:47163", "accept": false, - "reason": "NeuCodec model support vs paged-attention test additions; unrelated." + "reason": "NeuCodec model support vs paged-attention XPU tests; different features." }, { "left": "pull_request:46658", "right": "pull_request:47143", "accept": false, - "reason": "Step3.7 Flash support vs NeuCodec support; distinct model implementations." + "reason": "Step3.7 flash support and NeuCodec support are separate model integrations." }, { "left": "pull_request:46658", "right": "pull_request:47322", "accept": false, - "reason": "Step3.7 Flash support vs Quark config-version change; unrelated." + "reason": "Step3.7 flash conversion/model work vs Quark config-class changes; unrelated." }, { "left": "pull_request:47143", "right": "pull_request:47387", "accept": false, - "reason": "NeuCodec support vs one-to-many conversion plumbing; unrelated." + "reason": "NeuCodec model support vs one-to-many conversion infra; not the same change." }, { "left": "pull_request:47143", "right": "pull_request:47322", "accept": false, - "reason": "NeuCodec support vs Quark config-class change; unrelated." + "reason": "NeuCodec model support vs AMD Quark config changes; unrelated." }, { "left": "pull_request:45157", "right": "pull_request:47143", "accept": false, - "reason": "Prism GGUF support vs NeuCodec support; different model families and code paths." + "reason": "Prism GGUF dequantization and NeuCodec model support touch different subsystems." }, { "left": "pull_request:47301", "right": "pull_request:47367", "accept": false, - "reason": "MTP docs vs pipeline device-placement behavior; unrelated documentation topics." + "reason": "MTP docs vs pipeline device-placement clarification; different documentation topics and code semantics." }, { "left": "pull_request:47163", "right": "pull_request:47292", "accept": false, - "reason": "Different XPU test fixes for different model suites; not the same bug." + "reason": "Both are XPU test-related, but they target different models/features and different failures." }, { "left": "pull_request:46912", "right": "pull_request:47452", "accept": false, - "reason": "RoPE test coverage vs mamba coverage cleanup; unrelated test work." + "reason": "RoPE tests vs broad Mamba standardization/test coverage; similar testing theme only, not the same issue." }, { "left": "pull_request:47306", "right": "pull_request:47367", "accept": false, - "reason": "Neuron experiment scripts vs pipeline device-placement clarification; unrelated." + "reason": "Neuron experimentation scripts vs pipeline device placement; unrelated." }, { "left": "pull_request:40546", "right": "pull_request:43451", "accept": false, - "reason": "VibeVoice and Molmo2 are separate model additions, not one duplicate change." + "reason": "VibeVoice and Molmo2 are different model additions with separate docs, configs, and code." }, { "left": "pull_request:43451", "right": "pull_request:47143", "accept": false, - "reason": "Molmo2 vs NeuCodec are separate model support PRs." + "reason": "Molmo2 and NeuCodec are distinct model integrations." }, { "left": "pull_request:43451", - "right": "pull_request:47163", + "right": "pull_request:46509", "accept": false, - "reason": "Molmo2 model addition vs paged-attention XPU test work; unrelated." + "reason": "Molmo2 and Nemotron Omni are unrelated model-support PRs." }, { "left": "pull_request:43451", - "right": "pull_request:46509", + "right": "pull_request:47096", "accept": false, - "reason": "Molmo2 vs Nemotron Omni are distinct model-support PRs." + "reason": "Molmo2 model integration vs Kimi exportability work; different concrete changes." }, { "left": "pull_request:45273", "right": "pull_request:47298", "accept": false, - "reason": "Liger eval OOM fix vs MI450 model failures; unrelated." - }, - { - "left": "pull_request:43451", - "right": "pull_request:47096", - "accept": false, - "reason": "Molmo2 model support vs exportability changes for Kimi; different features." + "reason": "Liger eval OOM fix in trainer vs MI450 model failures; different runtime issues." }, { "left": "pull_request:43451", "right": "pull_request:47003", "accept": false, - "reason": "Molmo2 model support vs OpenVINO exporter work; unrelated." + "reason": "Molmo2 model support vs OpenVINO exporter support; different code paths and goals." }, { "left": "pull_request:47003", "right": "pull_request:47096", "accept": false, - "reason": "OpenVINO exporter vs Kimi exportability; different exporter paths and goals." + "reason": "OpenVINO exporter addition and exportability fixes for Kimi are related to export, but they are not the same concrete change." }, { - "left": "pull_request:43451", - "right": "pull_request:47322", + "left": "pull_request:47003", + "right": "pull_request:47306", "accept": false, - "reason": "Molmo2 model support vs Quark config-version update; unrelated." + "reason": "OpenVINO exporter work vs Neuron experiments; unrelated." }, { - "left": "pull_request:47096", - "right": "pull_request:47322", + "left": "pull_request:47003", + "right": "pull_request:47387", "accept": false, - "reason": "Kimi exportability vs Quark config changes; unrelated." + "reason": "OpenVINO exporter support vs one-to-many conversion support; different infrastructure changes." }, { - "left": "pull_request:43451", + "left": "pull_request:47096", "right": "pull_request:47306", "accept": false, - "reason": "Molmo2 model support vs Neuron experiment scripts; unrelated." + "reason": "Kimi exportability work vs Neuron experiments; unrelated." }, { "left": "pull_request:47096", - "right": "pull_request:47306", + "right": "pull_request:47387", "accept": false, - "reason": "Kimi exportability vs Neuron experiments; unrelated." + "reason": "Kimi exportability vs one-to-many conversion infrastructure; not the same code-path problem." }, { - "left": "pull_request:47003", - "right": "pull_request:47306", + "left": "pull_request:47096", + "right": "pull_request:47143", "accept": false, - "reason": "OpenVINO exporter vs Neuron experiments; unrelated." + "reason": "Kimi exportability work vs NeuCodec support; different model/feature areas." }, { "left": "pull_request:47096", - "right": "pull_request:47387", + "right": "pull_request:47163", "accept": false, - "reason": "Kimi exportability vs one-to-many conversion framework; different plumbing." + "reason": "Exportability work vs paged-attention XPU tests; unrelated." }, { "left": "pull_request:47003", - "right": "pull_request:47387", + "right": "pull_request:47322", "accept": false, - "reason": "OpenVINO exporter vs one-to-many conversion framework; unrelated." + "reason": "OpenVINO exporter support vs AMD Quark config changes; unrelated." }, { - "left": "pull_request:47096", - "right": "pull_request:47143", + "left": "pull_request:43451", + "right": "pull_request:47322", "accept": false, - "reason": "Kimi exportability vs NeuCodec support; unrelated." + "reason": "Molmo2 model integration vs Quark config-class changes; unrelated." }, { - "left": "pull_request:47096", - "right": "pull_request:47163", + "left": "pull_request:46658", + "right": "pull_request:47143", "accept": false, - "reason": "Kimi exportability vs paged-attention XPU tests; unrelated." + "reason": "Step3.7 flash conversion tooling and NeuCodec support are different model projects." } ] }, @@ -26486,7 +26439,7 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "5d479cabad5a77282e07d720bde1f6f8bad7fb4259c4eb9156c2bac1154ace23", + "prepared_review_unit_hash": "bc756db4e3e48839b65ec0e06cea7840a1f8d4b34ba32f8b8239c1d4da0dfe69", "component_budget": { "node_count": 2, "item_count": 2, @@ -26508,18 +26461,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster contains two unrelated open feature PRs: one adds TimesFM covariate forecasting, the other adds DINOv3 image-classification support. They do not appear to be the same underlying change, so this is not a duplicate cluster.", - "confidence": 0.98, + "summary": "These are two unrelated feature PRs: one adds covariate forecasting to TimesFM, the other adds DINOv3 image-classification support. The soft similarity is superficial, so this is not a duplicate pair.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "PR 40755 is the stronger representative of this cluster by diff scope and implementation depth; PR 41224 is a separate DINOv3 feature PR, not the same change.", "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "PR 40755 is the best PR candidate here because it has the broader, more self-contained implementation across model code, utilities, tests, and docs.", "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different model families and code paths: TimesFM covariate forecasting vs. DINOv3 image-classification support. They are not mergeable as one PR and do not fix the same bug/change." + "reason": "Reject: these touch different models and code paths (TimesFM covariates forecasting vs DINOv3 image classification support) and could not plausibly be merged into one PR." } ] }, @@ -26612,7 +26565,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "0eed2b096edeac62d618c3084fbf5d2d9497c701fd8936116fbb5d72c96b6657", + "prepared_review_unit_hash": "8e68dd5f635a158e23994eefb2fa240f800e5720cab57a2eccecd381910c6c17", "component_budget": { "node_count": 2, "item_count": 2, @@ -26634,18 +26587,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two separate GitHub Actions dependency upgrade PRs: one is a broad Node 24 compatibility pass across core workflow actions, the other is a Docker build action version bump. They are related in theme but not the same underlying change.", - "confidence": 0.94, + "summary": "These are related GitHub Actions dependency bumps, but they address different actions and different failure surfaces: Node 24 compatibility for core workflow actions vs. docker/build-push-action version upgrades.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the broader and more representative workflow-actions upgrade, centered on Node 24 compatibility across many shared workflows.", + "canonical_pr_reason": "PR 42975 is the broader, more representative workflow-actions upgrade: it updates checkout/cache/upload-artifact across many workflows and matches the cluster\u2019s overall Node 24 compatibility theme.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the best single representative of the cluster because it updates multiple widely used workflow actions and matches the shared upgrade theme more directly.", + "best_pr_reason": "PR 42975 is the best representative PR because it covers the wider set of workflow-action incompatibilities and is more likely to serve as the umbrella change for this cluster.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Both are workflow-maintenance PRs, but they change different action families and different code paths: core GitHub Actions runtime-related upgrades vs docker/build-push-action bumps. They are not the same concrete change and would not be merged into one PR." + "reason": "They are both workflow maintenance upgrades, but not the same concrete change: 42975 updates general GitHub Actions used for Node 24 compatibility, while 42976 only bumps docker/build-push-action. They could be merged independently and do not fix the same underlying bug." } ] }, @@ -26680,58 +26633,58 @@ "pull_request:43448|pull_request:45490", "pull_request:45490|pull_request:46562" ], - "prepared_review_unit_hash": "68b166709d6794c08ee012675ea8d5ad7a94f4ad557335e73e9a8c9ebb7405a2", + "prepared_review_unit_hash": "15beb60189fbedc3256ed1ec20bf0c63457d374f62a97fec4b56a3faf3a5ca92", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13047, - "estimated_input_tokens": 3262, - "estimated_eval_tokens": 6780 + "serialized_chars": 13043, + "estimated_input_tokens": 3261, + "estimated_eval_tokens": 6778 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 4, - "serialized_chars": 13047, - "estimated_input_tokens": 3262, - "estimated_eval_tokens": 6780 + "serialized_chars": 13043, + "estimated_input_tokens": 3261, + "estimated_eval_tokens": 6778 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs are all model-addition changes, but for different models and codepaths. The overlap is mostly shared boilerplate (docs, auto-mapping, repo checks), so none of the soft pairs look like true duplicates.", - "confidence": 0.98, + "summary": "All four items are independent model-addition pull requests with overlapping Transformers boilerplate, not duplicate fixes or the same concrete change. The shared files are mostly docs, auto-mapping, and repo-check registration, which is expected for new-model integrations.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "No canonical PR: ViT NEPA, Molmo, CTSM, and Step3p7 are distinct model additions with different implementation files and targets; shared edits are generic registration/docs scaffolding.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "No single PR is a good global representative for a duplicate cluster here, since the items are unrelated model additions rather than one underlying change.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model additions (ViT NEPA vs Molmo); shared auto/docs files are boilerplate, not the same change." + "reason": "ViT NEPA and Molmo are unrelated model additions with different model families, files, and implementation details. They only overlap on generic docs/auto-registration paths, which is not enough to treat them as the same PR." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Adds different models and codepaths; similarity is only broad Transformers registration/docs work." + "reason": "These add different models in different domains (vision vs. time series) and touch distinct codepaths. Shared Transformers boilerplate does not indicate the same underlying change." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo and CTSM are unrelated model implementations, so this is not the same underlying change." + "reason": "Molmo and CTSM are separate model integrations with different architectures and support files. Their common filenames reflect standard new-model wiring, not a shared bug or duplicate change." }, { "left": "pull_request:45490", "right": "pull_request:46562", "accept": false, - "reason": "CTSM and Step3p7 are separate model integrations with different source files and behavior; not mergeable as one duplicate PR." + "reason": "CTSM and Step3p7 are unrelated model additions. They share registration/documentation machinery, but the concrete model code, processing, and intended behavior are different, so they should not be merged as duplicates." } ] }, @@ -26780,130 +26733,130 @@ "pull_request:46539|pull_request:46922", "pull_request:46886|pull_request:46922" ], - "prepared_review_unit_hash": "1171908aa04278789bb7f2593642b109ccb58873bc5e1c9c013c1d11323c6171", + "prepared_review_unit_hash": "4969ddec6465eec7f8cac5a6531d9d3b9b1bd73ce01ecff971e748386ad0a7a1", "component_budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19604, - "estimated_input_tokens": 4901, - "estimated_eval_tokens": 10058 + "serialized_chars": 19581, + "estimated_input_tokens": 4896, + "estimated_eval_tokens": 10048 }, "budget": { "node_count": 7, "item_count": 7, "soft_pair_count": 15, - "serialized_chars": 19604, - "estimated_input_tokens": 4901, - "estimated_eval_tokens": 10058 + "serialized_chars": 19581, + "estimated_input_tokens": 4896, + "estimated_eval_tokens": 10048 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only the two YOLOS pull requests look like the same underlying fix; the rest are unrelated despite similar \u201cfix\u201d wording or shared test/subsystem areas.", - "confidence": 0.94, + "summary": "Almost all PRs are unrelated despite superficial test/subsystem overlap. The only clear duplicate pair is the two YOLOS device_map=\"auto\" fixes, which address the same device mismatch in the same model path.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46922 is the strongest representative of the YOLOS device-mismatch fix: it addresses the concrete forward-pass device placement bug directly and also includes the no-split-module safeguard for device_map='auto'.", + "canonical_pr_reason": "PR 46922 is the better canonical representative: it fixes the YOLOS device mismatch directly in forward() and also adds the no-split-module change, making it the more complete version of the same fix target.", "best_issue_reason": null, - "best_pr_reason": "PR #46922 is the best overall fit because it contains the most complete code-path fix for the YOLOS device mismatch, not just the supporting model-splitting change.", + "best_pr_reason": "PR 46922 is the strongest overall candidate because it most directly addresses the YOLOS device_map auto failure and subsumes the narrower 46886 change.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Whisper tokenizer decoding/test cleanup is unrelated to the CLIP/ViT test-only changes." + "reason": "Different models and different changes; one is a Whisper tokenizer/test fix, the other is unrelated test edits for CLIP and ViT." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper decoding logic and FP8 dequantization/model-loading fixes are different problems in different subsystems." + "reason": "Whisper decoding logic is unrelated to the FP8 dequant/loading changes in the other PR." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "Whisper tokenizer changes do not match the generation watermarking validation fix." + "reason": "Both mention Whisper-related code, but one is a tokenizer decode fix while the other is generation/watermarking validation and broader model changes; not the same bug." }, { "left": "pull_request:43291", "right": "pull_request:46922", "accept": false, - "reason": "Whisper decoding/test work is unrelated to the YOLOS device-mismatch bug." + "reason": "Whisper tokenizer/tests and YOLOS device placement are unrelated issues." }, { "left": "pull_request:43665", - "right": "pull_request:44815", + "right": "pull_request:46539", "accept": false, - "reason": "These are unrelated test edits versus a dequantization/runtime-loading fix." + "reason": "Both are test/CI-oriented, but they target different failures in different models and code paths." }, { "left": "pull_request:43665", - "right": "pull_request:46380", + "right": "pull_request:44815", "accept": false, - "reason": "The CLIP/ViT test assertions are not the same change as generation config validation for watermarking." + "reason": "Unrelated test assertions versus FP8 dequantization/loading internals." }, { "left": "pull_request:43665", - "right": "pull_request:46539", + "right": "pull_request:46380", "accept": false, - "reason": "Model test tweaks are not the same as the CI memory/device-map infrastructure fix." + "reason": "These touch different subsystems and different failure modes; the overlap is only that both include test updates." }, { "left": "pull_request:43665", "right": "pull_request:46922", "accept": false, - "reason": "These touch different models and different bug classes; only both are test-related at a very broad level." + "reason": "CLIP/ViT test edits are unrelated to the YOLOS device mismatch fix." }, { "left": "pull_request:44815", "right": "pull_request:46539", "accept": false, - "reason": "FP8 dequant/model-loading changes are unrelated to CI runner memory limits and test harness patching." + "reason": "FP8 kernel/loading fixes and Phimoe CI memory patch address different problems." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "The FP8 quantization path and generation/watermark validation are distinct code paths and bugs." + "reason": "Different code paths: FP8 dequantization versus generation config/watermarking and related model fixes." }, { "left": "pull_request:44815", "right": "pull_request:46922", "accept": false, - "reason": "FP8 dequant fixes and YOLOS device placement fixes are different underlying issues." + "reason": "No concrete overlap beyond both being model-related fixes; the bugs and code paths are different." }, { "left": "pull_request:46380", "right": "pull_request:46539", "accept": false, - "reason": "Generation config validation and CI integration-test environment setup are not the same bug/change." + "reason": "Generation/watermarking validation is unrelated to the Phimoe CI memory workaround." }, { "left": "pull_request:46380", "right": "pull_request:46922", "accept": false, - "reason": "Watermarking validation work is unrelated to the YOLOS device mismatch fix." + "reason": "Different models and failure modes; watermarking/config validation does not match the YOLOS device placement bug." }, { "left": "pull_request:46539", "right": "pull_request:46922", "accept": false, - "reason": "CI CPU-memory patching for integration tests is not the same as the YOLOS model runtime bug." + "reason": "CI runner memory limits for Phimoe are unrelated to YOLOS device_map auto behavior." }, { "left": "pull_request:46886", "right": "pull_request:46922", "accept": true, - "reason": "Both PRs target the same YOLOS device_map='auto' mismatch in the same file and issue; the later PR simply adds the direct tensor device cast on top of the same fix path." + "reason": "Both PRs fix the same YOLOS device_map auto device-mismatch problem in the same model file; 46922 simply includes a more complete version of the same fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "The judgments are grounded in the packet. The YOLOS pair (46886, 46922) is the only clearly mergeable same-bugfix pair because both target the same issue in the same file, and 46922 extends the earlier no-split-module fix with the direct device cast. The other pair rejections stay conservative and do not overstate similarity beyond soft similarity or broad subsystem overlap." + "feedback": "The summary is grounded in the packet, and the only accepted pair (46886 vs 46922) is a plausible duplicate/superset relationship for the same YOLOS device_map='auto' bugfix. The other soft-edge verdicts are conservative and do not overclaim mergeability or shared bug identity." }, "evaluator_used": true, "retried": false, @@ -26933,12 +26886,12 @@ "pull_request:43488|pull_request:43532", "pull_request:43532|pull_request:46747" ], - "prepared_review_unit_hash": "16ca0ca3d12a86d2883424d2458c4368a8b64716ecc8c0d6c92406f591fd3e72", + "prepared_review_unit_hash": "ed0126c680022c1a455e6eb01db4f17b7296a91426018407efa260e57645cf5e", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6350, + "serialized_chars": 6349, "estimated_input_tokens": 1588, "estimated_eval_tokens": 3432 }, @@ -26946,7 +26899,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 6350, + "serialized_chars": 6349, "estimated_input_tokens": 1588, "estimated_eval_tokens": 3432 }, @@ -26955,30 +26908,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster contains three unrelated '[don't merge]' PRs used for bot/workflow testing and repo checks. They do not fix the same underlying bug or implement the same change, despite superficial similarity in titles and some overlapping test/workflow surfaces.", - "confidence": 0.96, + "summary": "All three are separate \"[don't merge]\" test/CI PRs, but they change different workflows or setup code paths. The soft similarities are superficial; none looks like the same underlying bug or change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43532 is the most representative of the cluster because it touches the widest set of files and looks like the broadest exploratory change, but none of these PRs are true duplicates in the normal sense.", + "canonical_pr_reason": "No PR is a true canonical duplicate here; the cluster is a false-positive grouping of unrelated test/CI changes. If one must be picked as the most central, #43532 is the broadest workflow-centric change, but it is still not a duplicate of the others.", "best_issue_reason": null, - "best_pr_reason": "PR #43532 is the best single representative for triage only because it is the broadest and most central of the three test PRs; the others are narrower workflow toggles/checks.", + "best_pr_reason": "#43532 is the closest thing to a hub because it spans several CI/testing files, but its actual change set is still distinct from the setup/version tweak in #43488 and the security-gate workflow change in #46747.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:46747", "accept": false, - "reason": "Both are '[don't merge]' test PRs, but one changes packaging/versioning and Vit test files while the other only switches a workflow trigger for the security gate. Different code paths and different intended checks." + "reason": "Different areas and purposes: #43488 changes setup/dependency/version/test scaffolding, while #46747 only flips the security-gate workflow trigger. No shared code path or mergeable common fix." }, { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "Shared Vit test filename is incidental; #43488 is a setup/version-format repo-bot test, while #43532 rewrites multiple workflows and testing utilities. Not the same underlying change." + "reason": "The only overlap is a shared test file, but the actual edits are unrelated: #43488 updates setup/dependencies/version, while #43532 rewrites CI/workflow logic and related utilities. Not the same change." }, { "left": "pull_request:43532", "right": "pull_request:46747", "accept": false, - "reason": "#43532 is a broad workflow/test harness change; #46747 is a narrow PR CI security-gate trigger change. They affect different workflows and are not mergeable as one fix." + "reason": "Both touch GitHub Actions, but they modify different workflows with different behavior: self-comment/comment-driven CI versus the security-gate trigger. Too different to be one duplicate PR." } ] }, @@ -27008,7 +26961,7 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "5ef4ac788435b12fb88a2904156b8ff3754578a1b46ba3c63b6b71f544164e29", + "prepared_review_unit_hash": "fe65d6731ae2fb18789487ecfa3ea19d5ff38809231ab25dd3d39afc5d7711a2", "component_budget": { "node_count": 2, "item_count": 2, @@ -27030,18 +26983,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two related but distinct PRs: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. They share scaffolding and docs updates, but they target different features and code paths, so they should not be deduplicated.", + "summary": "These are related but distinct feature PRs: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. They share framework/docs scaffolding, but the actual pipelines, model hooks, and tests are for different code paths.", "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43613 is the broader and more complete visual-segmentation addition, with more model/test coverage and larger surface area; PR 43612 is a separate concept-segmentation feature, not a duplicate.", + "canonical_pr_reason": "No canonical PR: the two PRs implement different segmentation features and do not describe the same underlying change.", "best_issue_reason": null, - "best_pr_reason": "43613 is the better representative PR because it covers the more substantial pipeline implementation and associated tests/docs, while 43612 is a distinct feature request/implementation.", + "best_pr_reason": "PR 43613 is the broader standalone addition, but it is not a duplicate of 43612; both should be tracked separately.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different pipelines for different tasks (concept segmentation vs visual segmentation). Shared docs and registry edits are boilerplate, not evidence of the same underlying change." + "reason": "Different features and different pipeline modules/tests; shared docs/registry edits are incidental and do not indicate the same concrete change." } ] }, @@ -27071,40 +27024,40 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "5bfc49b0a9e7f271a8f8e5233a63767ba2acbafd329e2f37526d3eaf81910cab", + "prepared_review_unit_hash": "a2cce220e38e2991ac90bdcac6ab5fa5b5c2f9600dd08f746d4b05fdf8c4c8a3", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4293, - "estimated_input_tokens": 1074, - "estimated_eval_tokens": 2404 + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4293, - "estimated_input_tokens": 1074, - "estimated_eval_tokens": 2404 + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate PRs modifying the same file but addressing different trainer extensibility hooks: one for gradient-accumulation loss scaling, the other for signature-column filtering. They do not look like duplicates.", - "confidence": 0.97, + "summary": "Two open PRs in the same file, but they address different Trainer extension points: loss scaling for gradient accumulation vs. signature-column filtering for custom/PEFT trainers. They are related at a subsystem level, not the same underlying change.", + "confidence": 0.96, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "Neither PR is a duplicate of the other, so there is no canonical PR for this cluster. If one had to be highlighted, #43651 is a self-contained trainer behavior change around loss scaling, but it does not subsume #43663.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Different underlying changes: #43651 adds _loss_is_scaled_for_ga and changes training loss normalization; #43663 adds _get_signature_columns for column filtering. Same file, but not the same bug or mergeable fix." + "reason": "Both patch trainer extensibility in src/transformers/trainer.py, but one fixes gradient-accumulation loss scaling and adds _loss_is_scaled_for_ga, while the other extracts _get_signature_columns for input filtering. Different code paths and different behaviors, so not the same change." } ] }, @@ -27134,7 +27087,7 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "e53021fead7465aa6c47a0dd13688306e3a0d6485184189b99f6be526995aa18", + "prepared_review_unit_hash": "a329184fe5d044e39c1fd5eacfca763360fd9950165c48cc8121aa4d385794bc", "component_budget": { "node_count": 2, "item_count": 2, @@ -27156,18 +27109,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same CLI file and both mention TypeAdapter/name-resolution issues, but they fix different runtime problems: one is a broader annotation cleanup, the other is a pydantic-absent import/typing fix. They are related, not duplicates.", - "confidence": 0.89, + "summary": "Both PRs touch `src/transformers/cli/serve.py`, but they fix different `TypeAdapter`-related problems and should not be deduped.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43842 is the more focused and concrete fix: it directly addresses the TypeAdapter NameError when pydantic is unavailable by keeping the import in TYPE_CHECKING and quoting the annotation.", + "canonical_pr_reason": "PR #43842 is the more concrete and targeted fix: it explicitly guards the `TypeAdapter` import behind `TYPE_CHECKING` and quotes the annotation to avoid a runtime NameError when pydantic is absent.", "best_issue_reason": null, - "best_pr_reason": "PR #43842 is the best representative of the cluster because it targets a specific runtime failure mode with the smallest, clearest change set; PR #43656 looks like a separate annotation-related cleanup/fix.", + "best_pr_reason": "PR #43842 is the best representative of the cluster because it addresses a specific import-time failure mode with a narrowly scoped fix; PR #43656 is a broader typing cleanup that appears to target a different NameError path.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Not the same underlying bug. #43656 is a broader typing/future-annotations change in serve.py, while #43842 specifically fixes the TypeAdapter NameError caused by pydantic not being installed. They could coexist in one file but are not the same change." + "reason": "Same file and similar symptom, but the underlying bugs differ: #43656 is about general typing/annotation NameError cleanup, while #43842 specifically fixes `TypeAdapter` import handling when pydantic is not installed. They are not the same change and would not plausibly merge into one PR." } ] }, @@ -27300,8 +27253,8 @@ "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:44086|pull_request:44722", "pull_request:44018|pull_request:44019", "pull_request:44019|pull_request:44071", "pull_request:44068|pull_request:44116", @@ -27345,8 +27298,8 @@ "pull_request:44010|pull_request:44073", "pull_request:43997|pull_request:44074", "pull_request:44010|pull_request:44018", - "pull_request:43997|pull_request:44129", "pull_request:44071|pull_request:44129", + "pull_request:43997|pull_request:44129", "pull_request:43996|pull_request:44161", "pull_request:43996|pull_request:44017", "pull_request:44010|pull_request:44044", @@ -27360,268 +27313,244 @@ "pull_request:44044|pull_request:44154", "pull_request:44024|pull_request:44076" ], - "prepared_review_unit_hash": "c63573fa1baa29dfd95e63dd9da998995d988765b1c547546663ac07149c06b0", + "prepared_review_unit_hash": "950019d22b1e6489cd4472b85ce802ce62e326d8da485532cd4fe24bc27b1240", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84505, - "estimated_input_tokens": 21127, - "estimated_eval_tokens": 42510 + "serialized_chars": 84514, + "estimated_input_tokens": 21129, + "estimated_eval_tokens": 42514 }, "budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 72, - "serialized_chars": 76935, - "estimated_input_tokens": 19234, - "estimated_eval_tokens": 38724 + "serialized_chars": 76945, + "estimated_input_tokens": 19237, + "estimated_eval_tokens": 38730 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a broad standardized output-tracing refactor series across many different model families, not one duplicate cluster. The only clear duplicate-like pairs are the GPT-Neo variants and the GPT-J/CodeGen variants that touch the same concrete code paths.", - "confidence": 0.88, + "summary": "This cluster is mostly a loose theme of output-tracing refactors across many unrelated model families. The only clear duplicate groups are the GPT-J/CodeGen trio and the GPT-Neo pair; the rest share a tracking issue or broad subsystem but not the same concrete code change.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44722 is the strongest representative of the GPT-J/CodeGen output-tracing refactor slice: it targets the same concrete files and interface change as the accepted duplicates, and looks like the most complete/latest rewrite of that change.", + "canonical_pr_reason": "44066 is the clearest representative of the GPT-J/CodeGen refactor: explicit issue linkage, concrete code-path changes in both files, and a focused title matching the actual diff.", "best_issue_reason": null, - "best_pr_reason": "PR 44722 is the best overall PR representative here because it captures the core GPT-J/CodeGen decorator-based output-tracing migration with the clearest, most consolidated diff among the actually overlapping artifacts.", + "best_pr_reason": "44066 is the strongest single representative in the cluster because it is the most directly scoped, issue-linked, and concrete implementation of the standardized output-tracing change.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families; only a shared tracking issue, not the same code-path change." + "reason": "Different model families (CVT/FNet vs GPT-J/CodeGen); shared issue target alone is not a duplicate." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "CVT/FNet refactor vs DeBERTa refactor; same theme, different implementation and model code." + "reason": "CVT/FNet output-tracing refactor is unrelated to DeBERTa's attention/output handling." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Both are GPT-J output-tracing refactors on the same code path; 44066 just also bundles CodeGen." + "reason": "Same GPT-J/CodeGen output-tracing refactor on the same concrete code path; plausible to merge as one PR." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "ResNet/RT-DETR backbones vs EfficientNet; same issue target does not make them duplicates." + "reason": "Both are vision backbones, but they change different models and code paths; too broad to treat as duplicates." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet output-tracing refactor and GPT-J/CodeGen refactor are unrelated model implementations." + "reason": "EfficientNet standardized outputs and GPT-J/CodeGen tracing are unrelated code changes." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen vs EfficientNet; not the same underlying bug or change." + "reason": "GPT-J/CodeGen refactor vs EfficientNet refactor; same theme, different concrete change." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 refactor vs DeBERTa refactor; only broadly similar output-tracing work." + "reason": "MobileNetV2 and DeBERTa are unrelated implementations; only the issue link is common." }, { - "left": "pull_request:44066", - "right": "pull_request:44086", + "left": "pull_request:44086", + "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen refactor vs MGP-STR refactor; different modules and code paths." + "reason": "MGP-STR and MPT touch different model code paths, so they are not the same change." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen refactor vs MPT refactor; same theme, different concrete change." + "reason": "GPT-J/CodeGen output tracing and MPT output tracing are separate model-specific refactors." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo output-tracing migration touching the same file and interface; these look like duplicate variants of the same change." + "reason": "Both are GPT-Neo output-tracing refactors in the same file and same code path." }, { "left": "pull_request:44066", "right": "pull_request:44068", "accept": false, - "reason": "GPT-J/CodeGen refactor vs GPT-Neo refactor; different models and different code paths." + "reason": "GPT-J/CodeGen vs GPT-Neo; same abstraction, different model implementation." }, { - "left": "pull_request:44018", - "right": "pull_request:44066", + "left": "pull_request:44019", + "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo vs GPT-J/CodeGen; same refactor pattern but not the same artifact-level change." + "reason": "ResNet output tracing and GPT-J/CodeGen output tracing are not the same underlying change." }, { "left": "pull_request:44019", - "right": "pull_request:44722", + "right": "pull_request:44085", "accept": false, - "reason": "ResNet refactor and GPT-J/CodeGen refactor are unrelated despite both using output-tracing decorators." + "reason": "ResNet and GPT-J/CodeGen are unrelated model-specific refactors." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J/CodeGen are different architectures and code paths." - }, - { - "left": "pull_request:44019", - "right": "pull_request:44085", - "accept": false, - "reason": "ResNet vs GPT-J; no shared concrete change." + "reason": "MGP-STR and GPT-J/CodeGen are unrelated implementations." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet output tracing are separate model-specific refactors." + "reason": "GPT-Neo and ResNet are different concrete code paths." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet vs MPT; same subsystem theme only." + "reason": "ResNet output tracing and MPT output tracing are separate refactors in different models." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert are unrelated model implementations." + "reason": "GPT-Neo and Flaubert are unrelated model changes." }, { "left": "pull_request:44018", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo vs Flaubert; not the same underlying change." + "reason": "GPT-Neo and Flaubert do not share the same code path." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo refactor and GPT-J/CodeGen refactor are different models and diffs." + "reason": "GPT-Neo and GPT-J/CodeGen are different model families and different implementations." }, { "left": "pull_request:44071", "right": "pull_request:44722", "accept": false, - "reason": "MPT output handling and GPT-J/CodeGen output tracing are separate refactors." + "reason": "MPT output tracing and GPT-J/CodeGen output tracing are unrelated changes." }, { "left": "pull_request:44018", "right": "pull_request:44085", "accept": false, - "reason": "GPT-Neo and GPT-J are different code paths even if the refactor pattern is similar." + "reason": "GPT-Neo vs GPT-J/CodeGen; similar refactor pattern, but not the same PR." }, { "left": "pull_request:44018", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo vs MGP-STR; not a duplicate pair." + "reason": "GPT-Neo and MGP-STR are unrelated model-specific changes." }, { "left": "pull_request:44071", "right": "pull_request:44085", "accept": false, - "reason": "MPT and GPT-J are different models and changes." + "reason": "MPT and GPT-J/CodeGen differ in model and code path." }, { "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "GPT-Neo refactor vs MGP-STR refactor; only superficially similar." + "reason": "GPT-Neo and MGP-STR are different concrete implementations." }, { - "left": "pull_request:44085", + "left": "pull_request:44059", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J output tracing vs MGP-STR output tracing; different model code." + "reason": "GPT-2 output tracing and MGP-STR output tracing are separate model changes." }, { "left": "pull_request:44059", - "right": "pull_request:44086", + "right": "pull_request:44722", "accept": false, - "reason": "GPT2 refactor and MGP-STR refactor are unrelated changes." + "reason": "GPT-2 and GPT-J/CodeGen are different code paths; same refactor style is not enough." }, { "left": "pull_request:44098", "right": "pull_request:44101", "accept": false, - "reason": "ViLT and Flaubert are unrelated model-specific output-handling changes." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-Neo vs MPT; same general refactor theme, different concrete code." + "reason": "ViLT and Flaubert are unrelated model implementations." }, { "left": "pull_request:44017", "right": "pull_request:44073", "accept": false, - "reason": "Segformer and VisualBert are different model implementations." + "reason": "SegFormer and VisualBert are different model code paths." }, { "left": "pull_request:44017", "right": "pull_request:44161", "accept": false, - "reason": "Segformer and LongT5 are unrelated; shared issue target is too broad." - }, - { - "left": "pull_request:44073", - "right": "pull_request:44101", - "accept": false, - "reason": "VisualBert vs Flaubert; no shared code path." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44017", - "accept": false, - "reason": "RegNet and Segformer are distinct backbones with separate refactors." + "reason": "SegFormer and LongT5 are unrelated changes despite both using tracing decorators." }, { - "left": "pull_request:44101", - "right": "pull_request:44161", + "left": "pull_request:44059", + "right": "pull_request:44071", "accept": false, - "reason": "Flaubert and LongT5 are unrelated model changes." + "reason": "GPT-2 and MPT output tracing are different concrete implementations." }, { - "left": "pull_request:43997", - "right": "pull_request:44056", + "left": "pull_request:44056", + "right": "pull_request:44101", "accept": false, - "reason": "RegNet vs MPNet; different architectures and files." + "reason": "MPNet and Flaubert are unrelated model-specific refactors." }, { "left": "pull_request:44071", "right": "pull_request:44076", "accept": false, - "reason": "MPT and ImageGPT are separate model refactors." + "reason": "MPT and ImageGPT are different model families and different code paths." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet and Flaubert are unrelated implementations." + "reason": "TextNet and Flaubert are unrelated model changes." }, { "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "MPNet and LongT5 are different model-specific changes." + "reason": "MPNet and LongT5 are separate concrete changes." }, { "left": "pull_request:43997", "right": "pull_request:44018", "accept": false, - "reason": "RegNet and GPT-Neo are unrelated; only a shared output-tracing goal." + "reason": "RegNet and GPT-Neo are unrelated model implementations." }, { "left": "pull_request:43997", @@ -27633,133 +27562,115 @@ "left": "pull_request:44074", "right": "pull_request:44161", "accept": false, - "reason": "TextNet and LongT5 refactors are not the same artifact." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44161", - "accept": false, - "reason": "GPT-Neo and LongT5 are unrelated." + "reason": "TextNet and LongT5 are unrelated model-specific refactors." }, { - "left": "pull_request:44010", - "right": "pull_request:44071", + "left": "pull_request:43997", + "right": "pull_request:44056", "accept": false, - "reason": "SqueezeBert and MPT changes are different model-specific refactors." + "reason": "RegNet and MPNet are different models with different output-tracing logic." }, { - "left": "pull_request:44018", + "left": "pull_request:44071", "right": "pull_request:44129", "accept": false, - "reason": "GPT-Neo and SpeechT5 are unrelated code paths." + "reason": "MPT and SpeechT5 are unrelated refactors." }, { - "left": "pull_request:44010", - "right": "pull_request:44073", + "left": "pull_request:44074", + "right": "pull_request:44101", "accept": false, - "reason": "SqueezeBert and VisualBert are separate model implementations." + "reason": "TextNet and Flaubert are different concrete changes." }, { - "left": "pull_request:43997", - "right": "pull_request:44074", + "left": "pull_request:44056", + "right": "pull_request:44161", "accept": false, - "reason": "RegNet and TextNet are different models and diffs." + "reason": "MPNet and LongT5 are not the same underlying bug/change." }, { "left": "pull_request:44010", - "right": "pull_request:44018", - "accept": false, - "reason": "SqueezeBert and GPT-Neo are not duplicates." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44129", + "right": "pull_request:44071", "accept": false, - "reason": "RegNet and SpeechT5 are unrelated." + "reason": "SqueezeBert and MPT touch different code paths." }, { - "left": "pull_request:44071", + "left": "pull_request:44018", "right": "pull_request:44129", "accept": false, - "reason": "MPT and SpeechT5 are separate model changes." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44161", - "accept": false, - "reason": "CVT/FNet refactor and LongT5 refactor are different implementations." + "reason": "GPT-Neo and SpeechT5 are unrelated model implementations." }, { - "left": "pull_request:43996", - "right": "pull_request:44017", + "left": "pull_request:44010", + "right": "pull_request:44073", "accept": false, - "reason": "CVT/FNet and Segformer are unrelated model-specific changes." + "reason": "SqueezeBert and VisualBert are different model-specific refactors." }, { - "left": "pull_request:44010", + "left": "pull_request:44018", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBert and DeBERTa are different code paths." + "reason": "GPT-Neo and DeBERTa are unrelated code changes." }, { "left": "pull_request:44044", "right": "pull_request:44074", "accept": false, - "reason": "DeBERTa and TextNet are unrelated." + "reason": "DeBERTa and TextNet are separate output-tracing refactors." }, { "left": "pull_request:44066", "right": "pull_request:44722", "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor; same concrete files and same decorator migration." + "reason": "Same GPT-J/CodeGen standardized tracing change; the diffs target the same code path and could be merged into one PR." }, { "left": "pull_request:44085", "right": "pull_request:44722", "accept": true, - "reason": "Both target the GPT-J output-tracing change in the same code path; 44722 looks like the broader/latest variant." + "reason": "Both appear to be the GPT-J/CodeGen output-tracing refactor on the same files and same underlying change." }, { "left": "pull_request:44085", "right": "pull_request:44129", "accept": false, - "reason": "GPT-J refactor and SpeechT5 refactor are unrelated." + "reason": "GPT-J/CodeGen and SpeechT5 are unrelated model-specific changes." }, { "left": "pull_request:43996", "right": "pull_request:44007", "accept": false, - "reason": "CVT/FNet and ResNet/RT-DETR are different model refactors." + "reason": "CVT/FNet and ResNet are different model families; only the thematic issue is shared." }, { "left": "pull_request:43996", "right": "pull_request:44129", "accept": false, - "reason": "CVT/FNet and SpeechT5 are unrelated model families." + "reason": "CVT/FNet output-tracing refactor and SpeechT5 output-tracing refactor are separate changes." }, { "left": "pull_request:44044", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa and SpeechT5 are different underlying changes." + "reason": "DeBERTa and SpeechT5 are unrelated concrete code paths." }, { "left": "pull_request:44044", "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa and VITS are unrelated refactors." + "reason": "DeBERTa and Vits are different model implementations." }, { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "FocalNet and ImageGPT are different model-specific output-tracing refactors." + "reason": "FocalNet and ImageGPT are unrelated model-specific refactors; same decorator pattern is not enough." } ] }, "evaluator_result": { "accept": true, - "feedback": "The verdicts are conservative and grounded in the packet. The accepted pairs are limited to the clear same-file/same-code-path GPT-Neo and GPT-J/CodeGen variants, while the cross-family pairs are appropriately rejected. The summary also stays within the evidence and does not overclaim broader duplication." + "feedback": "The summary stays within the packet evidence: it identifies the strongest duplicate groupings conservatively and avoids claiming unrelated model refactors are duplicates. The accepted soft-edge pairs all look mergeable as the same output-tracing bugfix within the same model/file scope." }, "evaluator_used": true, "retried": false, @@ -27794,20 +27705,20 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "e0af67edc1e13a582629194b6eafb016d979a36d737cbe0d0a85d36af597fb7e", + "prepared_review_unit_hash": "a2541232114ce2c148d69200a43337730c2cd41f2805be6d604d06a4e4070565", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 77, - "serialized_chars": 84505, - "estimated_input_tokens": 21127, - "estimated_eval_tokens": 42510 + "serialized_chars": 84514, + "estimated_input_tokens": 21129, + "estimated_eval_tokens": 42514 }, "budget": { "node_count": 6, "item_count": 6, "soft_pair_count": 5, - "serialized_chars": 13070, + "serialized_chars": 13069, "estimated_input_tokens": 3268, "estimated_eval_tokens": 6792 }, @@ -27816,42 +27727,42 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "All items are model-specific refactors toward standardized output tracing/capture, but they affect different concrete code paths and should not be deduplicated together. No pair looks like the same underlying bug or change.", - "confidence": 0.91, + "summary": "All six PRs are model-specific refactors around standardized output tracing/capture_outputs, but they touch different architectures and code paths. They are related thematically, not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "SpeechT5 (PR 44129) is the strongest representative: it has the clearest scoped refactor, a larger output-tracing change, and an explicit issue target.", + "canonical_pr_reason": "No single canonical PR: each change targets a different model implementation (RegNet, SqueezeBert, MPNet, VisualBert, TextNet, SpeechT5) with separate output-tracing mechanics.", "best_issue_reason": null, - "best_pr_reason": "PR 44129 is the best single representative of the cluster because it is substantial, clearly scoped, and explicitly linked to issue 43979; the others are separate model migrations rather than duplicates.", + "best_pr_reason": "44129 is the strongest standalone representative because it has an explicit issue target and a clearly scoped SpeechT5 refactor, but it is still not a duplicate of the other model-specific PRs.", "soft_edge_verdicts": [ { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "Same refactor theme, but different model files and execution paths (VisualBert vs TextNet). They are not mergeable as one duplicate PR." + "reason": "Same refactor theme, but VisualBert and TextNet are different models with different forward/output paths; not the same underlying change." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Both standardize output tracing, but MPNet and VisualBert are unrelated model implementations with different code paths." + "reason": "MPNet vs VisualBert: distinct architectures and output-tracing implementations, so they should not be merged as one PR." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "MPNet and TextNet are separate model-specific refactors; the shared decorator pattern is too broad to treat as the same change." + "reason": "MPNet and TextNet are separate model code paths; shared capture_outputs usage is too broad to treat as duplicate work." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBert touch different model architectures and different forward logic; this is only a shared modernization pattern." + "reason": "RegNet and SqueezeBert are unrelated model implementations; this is a common refactor pattern, not the same bug/change." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBert and SpeechT5 are distinct code paths and model families. They do not fix the same concrete bug or change." + "reason": "SqueezeBert and SpeechT5 fix different model-specific output tracing behavior; the explicit issue link on 44129 does not make them duplicates." } ] }, @@ -27931,274 +27842,274 @@ "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030" ], - "prepared_review_unit_hash": "6e6f5e57c7adf5463bdfa1a349c6303c9d5c43ce30c811611c70aad3a7800b6c", + "prepared_review_unit_hash": "84a96b45dad6c1ef4d356fb353c60feb0df670c9b92ffc157dd96f4e5895db5f", "component_budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40190, - "estimated_input_tokens": 10048, - "estimated_eval_tokens": 20352 + "serialized_chars": 40180, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 }, "budget": { "node_count": 13, "item_count": 13, "soft_pair_count": 40, - "serialized_chars": 40190, - "estimated_input_tokens": 10048, - "estimated_eval_tokens": 20352 + "serialized_chars": 40180, + "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 in the same broad theme of output-tracing refactors, but they target different model-specific forward paths and helpers. They are similar in implementation style, not duplicates of the same underlying change.", - "confidence": 0.94, + "summary": "All items are model-specific output-tracing refactors. They share a common pattern and helper usage, but they target different architectures and code paths, so they are not true duplicates.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44025 is the clearest representative of the cluster: it is a self-contained output-tracing refactor and covers a closely related pair of files in the depth-anything family.", + "canonical_pr_reason": "PR 44003 is the broadest representative here: it spans both `falcon_mamba` and `mamba` and adds shared output-capture plumbing, so it best anchors the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 44025 is the best single representative because it is concrete, self-contained, and broadly illustrates the refactor pattern without being tangled with unrelated model families.", + "best_pr_reason": "44003 is the most central and infrastructure-like PR in the set, making it the best single representative even though the others are distinct changes.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different model families (`timm_backbone` vs `upernet`); same refactor pattern, but not the same code path." + "reason": "Different model families (`timm_backbone` vs `upernet`) and different forward paths; same tracing theme only." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Different models and forward implementations; only the output-tracing style is shared." + "reason": "`timm_backbone` and `vision_text_dual_encoder` are unrelated code paths; not the same bug/change." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "`timm_backbone` and `univnet` are unrelated model paths; not one concrete fix." + "reason": "Different architectures (`timm_backbone` vs `univnet`); only the refactor pattern is similar." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different subsystems (`timm_backbone` vs mamba/falcon_mamba`); same utility pattern, not the same bug." + "reason": "`timm_backbone` refactor vs `mamba/falcon_mamba` output tracing; separate model implementations and changes." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "`timm_backbone` and `rwkv` are separate model implementations; no shared underlying change." + "reason": "`timm_backbone` and `rwkv` are distinct model families with different concrete forward paths." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "`mobilenet_v1` and `univnet` are unrelated models; both refactor output handling, but not the same change." + "reason": "`mobilenet_v1` and `univnet` are unrelated models; same refactor motif, not the same change." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different model code paths (`mobilenet_v1` vs `upernet`); too broad to merge as one PR." + "reason": "`mobilenet_v1` vs `upernet` touches different architectures and output code paths." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "`mobilenet_v1` and `depth_anything` are distinct model refactors; shared helper usage is not enough." + "reason": "`mobilenet_v1` and `depth_anything` are different model families; not mergeable as one duplicate fix." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Different architectures (`mobilenet_v1` vs `vision_encoder_decoder`); same theme only." + "reason": "`mobilenet_v1` refactor is unrelated to the `vision_encoder_decoder` wrapper change." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "`mobilenet_v1` and `speech_encoder_decoder` touch different forward paths; not duplicate fixes." + "reason": "`mobilenet_v1` and `speech_encoder_decoder` are separate code paths; only the refactor style overlaps." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "`vision_text_dual_encoder` and `univnet` are unrelated; both add output-capturing refactors, but separately." + "reason": "`vision_text_dual_encoder` and `univnet` are different model families with different forward logic." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different model families and signatures; not the same concrete bug/change." + "reason": "`vision_text_dual_encoder` vs `upernet` is a different model-specific refactor, not a duplicate." }, { "left": "pull_request:44000", "right": "pull_request:44003", "accept": false, - "reason": "`vision_text_dual_encoder` vs `falcon_mamba/mamba` are different code paths and fixes." + "reason": "Different concrete implementations (`vision_text_dual_encoder` vs `mamba/falcon_mamba`), so not the same change." }, { "left": "pull_request:44000", "right": "pull_request:44025", "accept": false, - "reason": "`vision_text_dual_encoder` and `depth_anything` are separate model-specific refactors." + "reason": "`vision_text_dual_encoder` and `depth_anything` are unrelated models; same tracing theme only." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "`univnet` and `upernet` are unrelated models; only the refactor pattern overlaps." + "reason": "`univnet` and `upernet` are distinct architectures; not the same underlying bug or fix." }, { "left": "pull_request:44001", "right": "pull_request:44003", "accept": false, - "reason": "Different model stacks (`univnet` vs mamba/falcon_mamba`); not one fix." + "reason": "`univnet` vs `mamba/falcon_mamba` affects separate code paths and files." }, { "left": "pull_request:44001", "right": "pull_request:44025", "accept": false, - "reason": "`univnet` and `depth_anything` touch different forward paths; not mergeable as a single PR." + "reason": "`univnet` and `depth_anything` are unrelated model-specific changes." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "`upernet` and `rwkv` are distinct models; same refactor motif, different changes." + "reason": "`upernet` and `rwkv` are different models with separate output handling paths." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "`upernet` and `mamba/falcon_mamba` are separate code paths; not duplicates." + "reason": "`upernet` refactor is unrelated to the `mamba/falcon_mamba` output-tracing change." }, { "left": "pull_request:44003", "right": "pull_request:44028", "accept": false, - "reason": "`falcon_mamba/mamba` and `superpoint` are unrelated model implementations." + "reason": "`mamba/falcon_mamba` vs `superpoint` are different architectures and different concrete code paths." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "`mamba` family vs `codegen` are different concrete forward-path changes." + "reason": "`mamba` refactor and `codegen` refactor are separate model changes; not a single duplicate fix." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "`timm_backbone` and `codegen` are unrelated subsystems; only broad refactor similarity." + "reason": "`timm_backbone` and `codegen` are unrelated implementations; same helper usage does not make them duplicates." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "`univnet` and `codegen` address different output/return handling paths." + "reason": "`univnet` and `codegen` touch different code paths and model families." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "`upernet` and `codegen` are separate model refactors; not the same underlying issue." + "reason": "`upernet` and `codegen` are separate model-specific refactors." }, { "left": "pull_request:44000", "right": "pull_request:44004", "accept": false, - "reason": "`vision_text_dual_encoder` and `codegen` are distinct code paths and fixes." + "reason": "`vision_text_dual_encoder` and `codegen` are unrelated code paths; not the same underlying change." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "`depth_anything` and `vision_encoder_decoder` are different model families; not one duplicate change." + "reason": "`depth_anything`/`prompt_depth_anything` vs `vision_encoder_decoder` are different model families and forward paths." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "`depth_anything` vs `speech_encoder_decoder` is a broad similarity, not a shared fix." + "reason": "`depth_anything`/`prompt_depth_anything` is unrelated to `speech_encoder_decoder`; same refactor pattern only." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "`depth_anything` and `superpoint` are unrelated forward implementations." + "reason": "`depth_anything` and `superpoint` are separate vision models with different output shapes and code paths." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "`depth_anything` and `rwkv` are separate model-specific refactors." + "reason": "`depth_anything` refactor is unrelated to `rwkv`; different model family and implementation." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "`depth_anything` and `dpr` touch different model code paths; not the same bug." + "reason": "`depth_anything` and `dpr` are different model families; not a duplicate change." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "`vision_encoder_decoder` and `rwkv` are unrelated models; same utility refactor only." + "reason": "`vision_encoder_decoder` and `rwkv` have no shared concrete bug/change; only thematic similarity." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "`vision_encoder_decoder` and `speech_encoder_decoder` are similar wrappers but still distinct code paths." + "reason": "Both are encoder-decoder wrappers, but vision and speech wrappers are separate concrete code paths and the diffs are model-specific." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "`vision_encoder_decoder` and `superpoint` are not the same underlying change." + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated model implementations." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "`vision_encoder_decoder` and `dpr` are separate models with separate fixes." + "reason": "`vision_encoder_decoder` and `dpr` are different model families and output paths." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "`speech_encoder_decoder` and `superpoint` are unrelated; no shared concrete bug." + "reason": "`speech_encoder_decoder` and `superpoint` are different architectures; not the same duplicate fix." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "`speech_encoder_decoder` and `rwkv` are different code paths and changes." + "reason": "`speech_encoder_decoder` and `rwkv` are unrelated changes in different model families." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "`speech_encoder_decoder` and `dpr` are distinct model-specific refactors." + "reason": "`speech_encoder_decoder` and `dpr` are separate code paths; same helper refactor does not make them duplicates." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "`superpoint` and `rwkv` are unrelated model implementations." + "reason": "`superpoint` and `rwkv` are distinct models with different forward logic." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "`superpoint` and `dpr` are different forward paths; not duplicates." + "reason": "`superpoint` and `dpr` are different model families and not the same underlying issue." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "`rwkv` and `dpr` are separate models with different concrete changes." + "reason": "`rwkv` and `dpr` are unrelated model-specific refactors; no shared concrete fix." } ] }, @@ -28228,7 +28139,7 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "14844e8ed9ceefb5118331e9e2bd94d0d1c5791e07e41cbff485fba84321fd6e", + "prepared_review_unit_hash": "1d098fdbe3f0899206ce04e6326eff5483b087774580635c78e1795909b8b348", "component_budget": { "node_count": 2, "item_count": 2, @@ -28250,18 +28161,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are unrelated model-addition PRs: one adds PenguinVL, the other adds Voxtral TTS. The shared autos/docs scaffolding is incidental, not a duplicate code change.", - "confidence": 0.98, + "summary": "These are two unrelated model-addition PRs that only overlap on standard Transformers registration/docs plumbing. They do not appear to fix or implement the same underlying change, so the soft similarity link should be rejected.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45401 is the better representative artifact because it has an explicit issue target and a tighter, self-contained scope.", + "canonical_pr_reason": "PR 45401 is the clearer representative: it is scoped to a single model integration, has an explicit issue target, and includes the full implementation/test wiring for Voxtral TTS.", "best_issue_reason": null, - "best_pr_reason": "PR #45401 is the better cluster representative for triage since it is more clearly anchored to a tracked request and is easier to summarize as a standalone change.", + "best_pr_reason": "PR 45401 is the stronger standalone PR for this cluster because it is more focused and traceable, whereas PR 44662 is a separate PenguinVL model addition with no concrete relation to Voxtral TTS.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Reject: they implement different models and code paths (PenguinVL vision-language vs Voxtral TTS). Shared files are only auto-registry/docs boilerplate, not the same underlying fix or feature." + "reason": "Different model integrations (PenguinVL vs Voxtral TTS) with distinct codepaths and purposes; the shared files are generic model-registration/docs changes, not evidence of the same bug or change." } ] }, @@ -28291,12 +28202,12 @@ "soft_pairs": [ "pull_request:44979|pull_request:45363" ], - "prepared_review_unit_hash": "79a0fb5b6f07e64a830da0c92dde97598de473503e5f99757024241ff611059b", + "prepared_review_unit_hash": "7607ea219c1b6c691e71890acd34e29e3e51c9531efbd52eb116bf1026d89a56", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4408, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -28304,7 +28215,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4407, + "serialized_chars": 4408, "estimated_input_tokens": 1102, "estimated_eval_tokens": 2460 }, @@ -28313,18 +28224,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs are about fusion in Transformers, but they target different layers: #44979 introduces a generic Module Fusion API, while #45363 wires n-to-1 kernel fusion through KernelConfig and adds an example. They are related, but not duplicates.", - "confidence": 0.92, + "summary": "These are related fusion-feature PRs, but they are not the same change: one introduces a standalone Module Fusion API, while the other wires n-to-1 kernel fusion into `KernelConfig` and model loading. They overlap conceptually yet target different implementation paths and could not safely be merged as duplicates.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "#44979 is the more foundational, self-contained PR: it adds the core Module Fusion API and tests, whereas #45363 builds a specific KernelConfig-based fusion path on top of that broader concept.", + "canonical_pr_reason": "PR #45363 is the broader, more concrete implementation: it integrates kernel fusion through `KernelConfig`, model loading, and integration code rather than only adding a standalone API scaffold.", "best_issue_reason": null, - "best_pr_reason": "#44979 is the best representative for the cluster because it defines the main abstraction rather than one concrete application, so it is the broadest and most central artifact here.", + "best_pr_reason": "PR #45363 is the best representative for this cluster because it covers the end-to-end fusion workflow and has the more actionable code-path change; PR #44979 is more of an isolated API foundation.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "They share the fusion theme, but not the same concrete change: #44979 adds a new module fusion API, while #45363 implements n-to-1 kernel fusion plumbing plus an example. Different files, different code paths, and not plausibly one merged PR." + "reason": "Related feature area, but different scope and code paths: #44979 adds a new `module_fusion.py` API layer, while #45363 implements n-to-1 kernel fusion via `KernelConfig` and integration hooks. They are not the same concrete change and should not be merged as duplicates." } ] }, @@ -28354,7 +28265,7 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "7174049e89e655d3773828266c0526a7f48e360aec7db9eab6bfb8db57c817f6", + "prepared_review_unit_hash": "c29d237b5fbb5dad68de55612aa709de3af05dc728590f9b6db6c7b15cf13452", "component_budget": { "node_count": 2, "item_count": 2, @@ -28376,18 +28287,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests target the same Mistral regex patching AttributeError in `_patch_mistral_regex` and make the same code change in `tokenization_utils_tokenizers.py`; 45317 is the stronger canonical PR because it includes the fix plus test coverage.", + "summary": "Both pull requests fix the same AttributeError in `_patch_mistral_regex` by switching from `tokenizer.backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; the later PR adds tests and is the cleaner representative.", "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45317 is the better canonical PR: it fixes the same underlying AttributeError as 45086 but adds test coverage and appears to be the more complete, merged-together version of the change.", + "canonical_pr_reason": "PR 45317 is the stronger canonical PR: it contains the same core fix as 45086 plus test coverage, making it the more complete representative of this change.", "best_issue_reason": null, - "best_pr_reason": "PR 45317 is the best overall representative because it contains the same fix as 45086 and also includes a test file, making it the more complete implementation of the bug fix.", + "best_pr_reason": "PR 45317 is the best representative because it covers the full bug fix and validation, while 45086 is an earlier, smaller version of the same patch.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same concrete bug fix in the same code path, with nearly identical diff in `src/transformers/tokenization_utils_tokenizers.py`; 45317 simply extends it with tests, so these could plausibly be unified into one PR." + "reason": "Same underlying bug and identical code-path fix in `_patch_mistral_regex`; 45317 is just the more complete version with tests, so these are duplicate/near-duplicate PRs." } ] }, @@ -28417,7 +28328,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "4c4e52a3925d2cfc1b5001095385dbcd64a925a41a6d0bf8a59b66ffe5a650a6", + "prepared_review_unit_hash": "507fdc7919b9fad262e07c8662eeb1336ab5971226252e39d314797ab48aad35", "component_budget": { "node_count": 2, "item_count": 2, @@ -28439,18 +28350,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs concern NemotronH, but they are not the same change: #45115 is a narrower refactor/conversion-mapping update, while #45569 is a larger architecture split introducing NemotronHDense/NemotronHSparse and related docs/model wiring.", - "confidence": 0.92, + "summary": "Two open PRs touch Nemotron H, but they address different scopes: one is a narrow conversion-mapping/refactor change, the other is a broader model-implementation and docs split into dense/sparse variants. They are related but not duplicates.", + "confidence": 0.88, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR: #45115 and #45569 overlap on NemotronH files, but they target different changes and are not duplicates.", + "canonical_pr_reason": "PR #45569 is the better canonical PR because it is the broader, more complete Nemotron H change: it introduces the dense/sparse model packages, updates auto-mapping, docs, and tests, whereas #45115 is a narrower structural conversion/refactor patch.", "best_issue_reason": null, - "best_pr_reason": "#45569 is the best representative of the cluster because it is the broader, more complete NemotronH change set, covering the new dense/sparse model split and the surrounding documentation/model registration work.", + "best_pr_reason": "PR #45569 best represents the cluster since it covers the main user-facing/modeling change and associated documentation/test updates; #45115 appears to be a supporting conversion-mapping refactor rather than the primary implementation.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Shared NemotronH files do not imply duplication here; #45115 focuses on conversion/refactor mapping, while #45569 introduces separate dense/sparse implementations and docs. Different concrete code-path changes." + "reason": "They are not the same concrete code-path fix. #45115 adds checkpoint conversion key renames for NemotronH/Bamba compatibility, while #45569 implements the Nemotron H split into dense/sparse models with docs and tests. Overlap in files does not make them mergeable as one PR." } ] }, @@ -28480,7 +28391,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "866e5ab60161f76c3a675a7ddf7beafd8a2f8cd25cbc4d2315db9595cbe44eeb", + "prepared_review_unit_hash": "c3c69b77f4fd108d0ab2fc1d7cf9f0e91f0ec5ff69516b1ea95aee85ea387fa5", "component_budget": { "node_count": 2, "item_count": 2, @@ -28502,18 +28413,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are both typing-related PRs in transformers, but they address different changes: one is a broad type-checking sweep across many files, while the other adds typing protocols/helpers for specific model/PEFT integrations. Not duplicates.", - "confidence": 0.92, + "summary": "These are both typing-related PRs in transformers, but they are not the same concrete change. PR 45415 is a broad type-checking sweep plus pyproject ignore changes across many files, while PR 45425 adds specific typing helpers/protocols and targeted integration fixes. I would not merge them as duplicates.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45425 is the more focused, concrete change centered on shared typing helpers and model/PEFT integration, whereas PR 45415 is a wide repository type-checking pass.", + "canonical_pr_reason": "45425 is the more coherent, code-focused typing change set; 45415 is a much broader mechanical type-checking sweep with config noise across many files.", "best_issue_reason": null, - "best_pr_reason": "PR 45425 is the better representative of this cluster because it has a narrower, well-defined code change and clearer implementation scope.", + "best_pr_reason": "45425 is the better representative because it targets a specific typing problem space with tighter scope and clearer code-path intent.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "Both are typing-related, but they fix different things: 45415 is a broad type-checking cleanup across many modules, while 45425 adds specific typing aliases/protocols for modeling_utils and PEFT-related code. They are not the same underlying change and would not plausibly merge into one PR." + "reason": "Both are about typing, but they fix different things: 45415 is a repository-wide type-checking cleanup, while 45425 introduces typing protocol aliases and specific model/integration adjustments. Shared filenames alone are not enough to treat them as the same PR." } ] }, @@ -28543,12 +28454,12 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "882da9332d7cad2a6e70eb0990d4b48c9431c5f4e2cfccdb8d005d7d96677229", + "prepared_review_unit_hash": "47c14f91fe721fafb8cf69f302956f4275635313a0cd7b6c493d497a15a931de", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4508, + "serialized_chars": 4507, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -28556,7 +28467,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4508, + "serialized_chars": 4507, "estimated_input_tokens": 1127, "estimated_eval_tokens": 2510 }, @@ -28565,18 +28476,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related expert-parallelism PRs, but they fix different bugs in different loading paths: 45548 targets EP + DeepSpeed ZeRO-3 / accelerate launch config handling, while 45662 targets EP + FSDP2 expert parameters being overwritten by rank-0 broadcast. Not duplicates.", - "confidence": 0.96, + "summary": "Both PRs are EP-related, but they fix different bugs: one addresses EP/DeepSpeed ZeRO-3 loading and config serialization, while the other fixes expert tensors being overwritten during FSDP2 rank-0 broadcast. Shared files alone aren\u2019t enough to treat them as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "pull_request:45662 is the better representative because it targets a concrete, well-scoped correctness bug (experts being overwritten by rank-0 broadcast) with a clearer end-to-end failure mode and broader EP/FSDP2 handling changes.", "best_issue_reason": null, - "best_pr_reason": "PR 45662 is the broader EP fix and the more representative PR for the cluster, but it still addresses a different concrete failure mode than 45548.", + "best_pr_reason": "pull_request:45662 is the strongest standalone PR in this cluster: it fixes a specific data corruption issue in distributed EP loading, touches the relevant loading and MoE/tensor-parallel paths, and is more clearly a single mergeable change than the other PR.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Different concrete code-path bugs: one is config/serialization for EP loading via accelerate + ZeRO-3, the other is FSDP2 expert parameters being overwritten during broadcast. Shared EP context is not enough to merge them." + "reason": "They share the EP/distributed-loading theme, but the underlying bugs differ: 45548 is about EP + DeepSpeed ZeRO-3 loading and stripping distributed_config from serialization, while 45662 fixes experts being silently overwritten by rank-0 broadcast under FSDP2. Not the same code-path problem." } ] }, @@ -28606,40 +28517,40 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "09d68dcb375c4beca71dea1e3e143eb9b251e294dc60f5ce83c623ad7e8c7acb", + "prepared_review_unit_hash": "df5ebd1a64c5b46f7663b64b4779686c7300f32a1fdba068fa01687084ead0b1", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4672, - "estimated_input_tokens": 1168, - "estimated_eval_tokens": 2592 + "serialized_chars": 4673, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4672, - "estimated_input_tokens": 1168, - "estimated_eval_tokens": 2592 + "serialized_chars": 4673, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs address the same xdist-related captured_info collision in testing/debug logging. PR 45645 is the broader, more complete fix because it includes the source-side file split plus CI workflow support for reading multiple captured_info files; PR 45639 is a narrower subset of the same change.", + "summary": "Both PRs target the same xdist collision around `captured_info` debug artifacts in the testing utilities. PR 45645 is the broader follow-up/superset: it includes the core file-name fix plus CI workflow updates to collect the new per-worker logs, while 45639 is the narrower core fix alone.", "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the better canonical PR because it covers the same underlying xdist-safe logging fix as 45639 and also updates CI to preserve/aggregate the resulting split debug logs, making it the more complete implementation.", + "canonical_pr_reason": "PR 45645 is the best canonical representative because it contains the underlying xdist-safe logging fix and the necessary CI artifact handling changes, making it the more complete version of the change.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 is the best representative of the cluster since it fixes the concrete captured_info collision and includes the associated CI handling needed for the new per-worker log files.", + "best_pr_reason": "PR 45645 best captures the full user-facing change set: fixing the collision in `testing_utils.py` and updating workflow/log collection so the new files are actually preserved in CI.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Accept: both PRs fix the same concrete xdist collision around captured_info output in testing_utils, and 45645 appears to be a superset that adds workflow support for the same change." + "reason": "Both address the same concrete bug: `captured_info` writes colliding under xdist. The second PR is a superset that adds workflow support for the same per-worker log files, so they could plausibly be merged into one PR." } ] }, @@ -28735,12 +28646,12 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "b1ecefaad87cc6f313dca6f29f3325d61ebf80844d34491aa56a3cbf6d76a6d3", + "prepared_review_unit_hash": "d2c60d37ab43a6afb03b2a42156dad5512192a9239d16de2abb67386103645b0", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5526, + "serialized_chars": 5525, "estimated_input_tokens": 1382, "estimated_eval_tokens": 3020 }, @@ -28748,7 +28659,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5526, + "serialized_chars": 5525, "estimated_input_tokens": 1382, "estimated_eval_tokens": 3020 }, @@ -28757,18 +28668,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are related through the distributed stack, but they do not look like the same underlying change. PR #46126 is a focused EP + Trainer integration, while PR #46269 is a broad distributed branch base spanning many files and subsystems.", - "confidence": 0.88, + "summary": "These PRs are related distributed-work items, but they are not the same change: #46269 is a broad distributed branch base, while #46126 is a narrower EP + Trainer integration built on top of that foundation. The shared filenames look like branch overlap rather than duplicate intent.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46126 is the more specific, self-contained change with a clear feature goal, making it a better canonical representative than the broad branch-base PR #46269.", + "canonical_pr_reason": "#46269 is the broader umbrella PR introducing the distributed branch base and multiple infra changes, so it is the better canonical representative of the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR #46126 is the clearest representative of this cluster because its title and diff point to one concrete feature area; #46269 is too broad and mixed in scope to serve as the best canonical PR.", + "best_pr_reason": "#46269 is the best representative because it subsumes the shared distributed groundwork; #46126 is a specific follow-on feature PR, not a duplicate of the same concrete fix.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Shared filenames only show overlap in the distributed subsystem. The titles and diffs indicate different scopes and likely different code-path changes: #46126 adds EP/router and Trainer integration, while #46269 is a large distributed branch base touching CI, FSDP, integrations, and model code. They are not the same fix and would not plausibly merge into one PR." + "reason": "Different scope and intent: #46126 adds EP/Trainer integration details, while #46269 establishes the distributed branch base. Shared files alone are not enough to treat them as the same PR." } ] }, @@ -28836,7 +28747,7 @@ "pull_request:46504|pull_request:46716", "pull_request:46432|pull_request:46716" ], - "prepared_review_unit_hash": "0fbef2d17efb55870ff08aa7badcc831bcb9048596e5b1795cf585a226a85908", + "prepared_review_unit_hash": "4877a81918c0d5ff80442ad8c639c66609c5acf868eee8e3bbc7232881f4bfa3", "component_budget": { "node_count": 10, "item_count": 10, @@ -28858,210 +28769,198 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are mostly unrelated workflow-maintenance PRs. The pr-ci-security-gate edits form a loose thematic cluster, but each PR changes a different control or permission aspect, so I do not see true duplicate PRs. The remaining PRs touch different workflows and are clearly separate.", + "summary": "This is a heterogeneous cluster of workflow-maintenance pull requests. The only apparent subcluster is pr-ci-security-gate.yml, but those PRs make different security changes and are not duplicate fixes. I would not merge any of the soft edges.", "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "No PR is a strict duplicate of the others. If a representative is needed, PR 46554 is the most substantive and concrete security-gate workflow change, but it still addresses a different mechanism than the other PRs.", + "canonical_pr_reason": "Best representative is PR 46554: it makes the most substantive functional change to the shared PR CI security gate, while the rest are either unrelated workflow tweaks or narrower follow-up edits to different concerns.", "best_issue_reason": null, - "best_pr_reason": "PR 46554 is the strongest representative of the cluster because it makes the most direct functional change to the pr-ci-security-gate workflow, rather than only tweaking permissions, comments, or unrelated workflow plumbing.", + "best_pr_reason": "PR 46554 is the strongest standalone representative because it changes the core security-gate logic rather than a peripheral workflow setting.", "soft_edge_verdicts": [ { "left": "pull_request:46138", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different changes: self-comment CI message piping vs PR CI security-gate logic." + "reason": "Different workflows and different problems: self-comment CI stdin handling vs PR CI security-gate hardening." }, { "left": "pull_request:46138", "right": "pull_request:46558", "accept": false, - "reason": "Different workflows with unrelated behavior; one updates self-comment CI, the other changes PR CI gate permissions/approval flow." + "reason": "Unrelated files and fixes; one updates self-comment CI, the other adjusts PR CI gate permissions/flow." }, { "left": "pull_request:46138", "right": "pull_request:46563", "accept": false, - "reason": "Unrelated workflows; one updates self-comment CI, the other adjusts a reusable security-gate workflow reference/comments." + "reason": "Different workflow files and different changes; not the same underlying bug or change." }, { "left": "pull_request:46138", "right": "pull_request:46716", "accept": false, - "reason": "Different workflow files and different concerns; no shared concrete bug or change." + "reason": "Self-comment CI command invocation is unrelated to the PR CI security gate secrets change." }, { "left": "pull_request:46138", "right": "pull_request:46504", "accept": false, - "reason": "Different workflows and different code paths; self-comment CI vs build-ci-docker-images input handling." + "reason": "Different workflows and behavior; one pipes comments to a helper script, the other changes CI image tagging logic." }, { "left": "pull_request:46432", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and unrelated changes; ssh-runner hardening vs benchmark dispatch trigger." + "reason": "SSH runner hardening is unrelated to benchmark workflow trigger changes." }, { "left": "pull_request:46432", "right": "pull_request:46501", "accept": false, - "reason": "Different workflow files and unrelated changes; ssh-runner hardening vs anti-slop permissions move." + "reason": "Different workflows and different fixes; SSH runner secret handling vs anti-slop permissions scoping." }, { "left": "pull_request:46432", "right": "pull_request:46502", "accept": false, - "reason": "Different workflow files and unrelated behavior; ssh-runner changes do not match reviewer-assignment ref selection." + "reason": "Unrelated workflow files; SSH runner secret cleanup is not the same change as reviewer assignment checkout ref selection." }, { "left": "pull_request:46432", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different goals; ssh-runner secret handling vs PR CI security-gate logic." + "reason": "Both touch CI security themes, but they fix different workflows and different code paths." }, { "left": "pull_request:46432", "right": "pull_request:46558", "accept": false, - "reason": "Different workflows and different fixes; ssh-runner secret handling vs PR CI gate permission/approval changes." - }, - { - "left": "pull_request:46432", - "right": "pull_request:46563", - "accept": false, - "reason": "Different workflows and different change types; ssh-runner hardening vs PR CI security-gate wrapper comments/reference." - }, - { - "left": "pull_request:46432", - "right": "pull_request:46716", - "accept": false, - "reason": "Different workflows and unrelated concerns; ssh-runner secret handling vs PR CI gate secrets forwarding." + "reason": "Different workflow files and different security concerns; not duplicate changes." }, { "left": "pull_request:46501", "right": "pull_request:46503", "accept": false, - "reason": "Same broad CI area but different workflows and different concrete behavior; anti-slop permissions vs benchmark dispatch trigger." + "reason": "Anti-slop permissions scoping and benchmark dispatch support are unrelated workflow changes." }, { "left": "pull_request:46501", "right": "pull_request:46502", "accept": false, - "reason": "Different workflow files and unrelated behavior; anti-slop permissions vs reviewer checkout ref." + "reason": "Different workflow jobs with different behavior; not the same underlying change." }, { "left": "pull_request:46501", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and different code paths; anti-slop permissions vs build-ci-docker-images commit-message handling." + "reason": "Anti-slop workflow permissions and build-ci image selection are separate concerns." }, { "left": "pull_request:46501", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different fixes; anti-slop permissions vs PR CI security-gate checks." + "reason": "Same broad CI area, but different workflows and different fixes." }, { "left": "pull_request:46502", "right": "pull_request:46503", "accept": false, - "reason": "Different workflow files and distinct CI behaviors; reviewer checkout ref vs benchmark dispatch support." + "reason": "Reviewer assignment checkout ref and benchmark dispatch support are unrelated." }, { "left": "pull_request:46502", "right": "pull_request:46504", "accept": false, - "reason": "Different workflow files and unrelated changes; reviewer assignment vs build image trigger parsing." + "reason": "Different workflows and different behavior; not duplicate." }, { "left": "pull_request:46502", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different problems; reviewer checkout ref vs PR CI security-gate logic." + "reason": "Different workflow files and different problems." }, { "left": "pull_request:46503", "right": "pull_request:46504", "accept": false, - "reason": "Different workflows and unrelated behavior; benchmark dispatch trigger vs build-ci-docker-images message parsing." + "reason": "Benchmark trigger expansion and build-ci commit-message parsing are unrelated changes." }, { "left": "pull_request:46503", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and different CI mechanisms; benchmark dispatch vs PR CI gate security checks." + "reason": "Different workflows and different code paths; no duplicate bug/change here." }, { "left": "pull_request:46504", "right": "pull_request:46554", "accept": false, - "reason": "Different workflows and distinct concerns; build-ci-docker-images parsing vs PR CI gate validation." + "reason": "Build-ci image tagging logic and PR CI security-gate logic are separate changes." }, { "left": "pull_request:46504", "right": "pull_request:46558", "accept": false, - "reason": "Different workflows and unrelated behavior; build-ci-docker-images parsing vs PR CI gate permissions/approval." + "reason": "Different workflows and different fixes; not the same underlying issue." }, { "left": "pull_request:46554", "right": "pull_request:46558", "accept": false, - "reason": "Same file, but different concrete changes: one adds checkout/diff-based security checks, the other changes permissions and approval gating." + "reason": "Same file, but different security hardening edits in different sections; they are related, not duplicates." }, { "left": "pull_request:46554", "right": "pull_request:46563", "accept": false, - "reason": "Same file, but different concrete changes: one implements security checks, the other only switches to a reusable workflow and adds explanatory comments." + "reason": "Both are PR CI gate edits, but one adds checks while the other only annotates the reusable workflow; not the same change." }, { "left": "pull_request:46554", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but different concrete changes: one adds security checks, the other changes secrets inheritance to an empty secrets map." + "reason": "Different modifications to the same workflow file; one adds gate logic, the other changes secret inheritance." }, { "left": "pull_request:46558", "right": "pull_request:46563", "accept": false, - "reason": "Same workflow file, but the edits target different concerns: permissions/approval flow vs reusable-workflow wrapper comments." + "reason": "Same workflow file but different concerns: permissions/approval gating vs reusable-workflow commentary." }, { "left": "pull_request:46558", "right": "pull_request:46716", "accept": false, - "reason": "Same workflow file, but different behavior changes: permission/approval gating vs secrets handling." + "reason": "Different security-gate edits in the same file, but not the same underlying fix." }, { "left": "pull_request:46563", "right": "pull_request:46716", "accept": false, - "reason": "Same file, but different changes: reusable workflow safety comments vs secrets inheritance removal." + "reason": "One adds safety notes, the other changes secrets handling; not duplicate." }, { "left": "pull_request:46504", "right": "pull_request:46563", "accept": false, - "reason": "Different workflows and unrelated change sets; build-image trigger parsing vs PR CI gate wrapper comments." + "reason": "Different workflow files and different fixes; no shared underlying change." }, { "left": "pull_request:46432", "right": "pull_request:46563", "accept": false, - "reason": "Different workflows and distinct security/automation concerns; ssh-runner hardening vs PR CI security-gate wrapper comments." + "reason": "SSH runner changes are unrelated to the reusable workflow note update in PR CI." }, { "left": "pull_request:46504", "right": "pull_request:46716", "accept": false, - "reason": "Different workflows and unrelated behavior; build-ci-docker-images parsing vs PR CI gate secrets handling." + "reason": "Build-ci docker image workflow changes are unrelated to PR CI secrets handling." }, { "left": "pull_request:46432", "right": "pull_request:46716", "accept": false, - "reason": "Different workflows and distinct concerns; ssh-runner secret handling vs PR CI gate secrets forwarding." + "reason": "Different workflows and different concerns; not a duplicate." } ] }, @@ -29154,7 +29053,7 @@ "soft_pairs": [ "pull_request:46310|pull_request:47162" ], - "prepared_review_unit_hash": "d69b8cc667755b2659b1c1ac859bb5b8ffb6abc3d71fe4633aadc73ebdc1c39d", + "prepared_review_unit_hash": "2eb7a3c7eccb3a26c3f3331517d0371cd384db684a9df6d600ac2c37458f9db0", "component_budget": { "node_count": 2, "item_count": 2, @@ -29176,18 +29075,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs fix the same off-by-one error in MoonshineStreaming\u2019s right-side sliding window mask (`-dist < right_window_size` to `<=`). PR #47162 is the stronger canonical choice because it includes the same MoonshineStreaming fix plus a test and an additional identical fix in Gemma4.", + "summary": "Both PRs fix the same off-by-one in MoonshineStreaming\u2019s sliding_window_mask_function right window (`<` to `<=`) and target the same issue, with 47162 also adding a test and an extra same-pattern fix in Gemma4.", "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #47162 is the best canonical PR: it covers the MoonshineStreaming off-by-one fix, adds test coverage, and is the more complete implementation of the same bug pattern.", + "canonical_pr_reason": "47162 is the better canonical PR because it contains the same MoonshineStreaming fix as 46310, adds test coverage, and appears to be the more complete follow-up.", "best_issue_reason": null, - "best_pr_reason": "PR #47162 is the best overall PR because it contains the same core fix as #46310 and is more complete due to the added test and related Gemma4 patch.", + "best_pr_reason": "47162 is the strongest representative of the cluster: it fixes the same MoonshineStreaming bug, includes validation, and is the more comprehensive change.", "soft_edge_verdicts": [ { "left": "pull_request:46310", "right": "pull_request:47162", "accept": true, - "reason": "Both PRs fix the same concrete off-by-one bug in MoonshineStreaming\u2019s right window mask, with the same code change in the shared files; they are plausibly mergeable as one change." + "reason": "Same concrete off-by-one bug in MoonshineStreaming sliding_window_mask_function, same target issue, and the diffs overlap on the exact code path; they could plausibly be merged into one PR." } ] }, @@ -29217,7 +29116,7 @@ "soft_pairs": [ "pull_request:46574|pull_request:46580" ], - "prepared_review_unit_hash": "ab4cbaabefc98dc55cded9d3430ad354441015939c4049a29660ebd27abea3fb", + "prepared_review_unit_hash": "858dd56a959c4004399061a39a41e1b2f455c432b60ad42527ad5f68c1dcf128", "component_budget": { "node_count": 2, "item_count": 2, @@ -29239,18 +29138,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated pull requests: one fixes CodeLlama decode stripping leading spaces, and the other fixes DeepSeek-Coder/DeepSeek-LLM loading the wrong tokenizer class. They should not be clustered as the same change.", - "confidence": 0.98, + "summary": "These are two distinct tokenizer-related fixes: one corrects CodeLlama decode stripping behavior, the other fixes DeepSeek-Coder/DeepSeek-LLM loading under the wrong tokenizer class. They share a broad area but not the same bug or code path.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46574 is the more self-contained bug fix: it changes CodeLlama tokenizer decoder setup to preserve real leading spaces on decode and includes targeted tests.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #46574 is the strongest standalone fix in this cluster because it directly corrects a concrete code path with minimal scope and clear test coverage.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:46574", "right": "pull_request:46580", "accept": false, - "reason": "Different underlying problems and code paths: CodeLlama decode-space stripping versus DeepSeek checkpoint-to-tokenizer-class resolution. They are not the same fix and would not plausibly merge into one PR." + "reason": "Different underlying problems and code paths: 46574 changes CodeLlama decoder setup to preserve leading spaces, while 46580 adds checkpoint-prefix routing in AutoTokenizer for specific DeepSeek models. They should not be merged as duplicates." } ] }, @@ -29283,7 +29182,7 @@ "pull_request:46903|pull_request:46958", "pull_request:46958|pull_request:47090" ], - "prepared_review_unit_hash": "6c60ad50199397d1ce0c06178b7135b51813e8ab000f5945f6954ff74adaece9", + "prepared_review_unit_hash": "d4bf563404603c5a7b0fa8d638e628c891936835c2fc59bf154bbf5428416778", "component_budget": { "node_count": 3, "item_count": 3, @@ -29305,36 +29204,39 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two PRs are the same CohereASR training-loss double-shift fix (#46958 and #47090); the PPFormulaNet/CohereASR bundle (#46903) is related but broader and not a clean duplicate of the others.", - "confidence": 0.83, + "summary": "These PRs are related follow-ups around training-loss double-shift handling, but they do not read like one clean duplicate cluster: 46903 also covers PPFormulaNet, 46958 is a CohereASR-only CrossEntropyLoss rewrite, and 47090 is a broader later loss-plumbing change that adds Moonshine variants. Treat them as overlapping fixes rather than one clearly mergeable single PR set.", + "confidence": 0.63, "canonical_issue_reason": null, - "canonical_pr_reason": "#47090 is the most representative fix for the shared CohereASR double-shift bug: it targets the same issue as #46958, has the broader and more mature patch set, and includes the same core code-path change rather than just a narrow local tweak.", + "canonical_pr_reason": "47090 is the strongest canonical PR because it covers the broadest affected code paths, keeps the loss plumbing aware of batch-aware reduction, and includes the most complete test coverage for the expanded model set.", "best_issue_reason": null, - "best_pr_reason": "#47090 is the best single PR to keep as canonical because it subsumes the CohereASR fix and extends it to additional affected models/tests while staying centered on the same loss-shift bug.", + "best_pr_reason": "47090 is the best standalone PR to keep: it addresses the shared loss-path problem with the most complete scope and test coverage across CohereASR and the later related variants.", "soft_edge_verdicts": [ { "left": "pull_request:46903", "right": "pull_request:47090", "accept": false, - "reason": "Both mention loss double-shift, but #46903 is a separate multi-model bundle centered on PPFormulaNet/issue 46901, while #47090 is a CohereASR-focused fix tied to issue 46894 and broader model set. Related, but not the same concrete PR change." + "reason": "Related loss-bug family, but not clearly the same mergeable patch: 46903 bundles CohereASR plus PPFormulaNet, while 47090 revises the CohereASR loss path again and expands to Moonshine variants." }, { "left": "pull_request:46903", "right": "pull_request:46958", "accept": false, - "reason": "They overlap on CohereASR, but #46903 also changes PPFormulaNet and is anchored to a different issue. The overlap looks like a shared pattern, not one duplicate change set." + "reason": "Both touch CohereASR double-shift handling, but 46903 is a broader multi-model fix and 46958 is a different CE-only implementation, so they do not look like one clearly mergeable PR." }, { "left": "pull_request:46958", "right": "pull_request:47090", - "accept": true, - "reason": "Same underlying CohereASR training-loss double-shift bug, same explicit issue target 46894, and the diffs fix the same code path; #47090 is a broader follow-up that could plausibly be merged into the same PR." + "accept": false, + "reason": "They are in the same bug family, but 47090 is a later broader loss-plumbing change with additional model coverage rather than the same concrete patch as 46958." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The packet supports a related family of double-shift fixes, but the summary overstates how confidently they are the same bug and mergeable as one PR set. PR 46903 is tied to a different issue target and covers CohereASR plus PPFormulaNet; PR 46958 is a separate CohereASR-only implementation using CrossEntropyLoss; PR 47090 is a later broader patch that changes the loss plumbing again and adds Moonshine variants. That looks more like overlapping follow-up fixes than one clearly mergeable single-bug PR cluster." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -29358,7 +29260,7 @@ "soft_pairs": [ "pull_request:47140|pull_request:47303" ], - "prepared_review_unit_hash": "32f13c6f483ed370000cf0030e98db75c52a08b09d51a77e898eeefab71741bf", + "prepared_review_unit_hash": "62992d23e77e028081d13227993c179e1b427a613697cae38db27698c2d4073c", "component_budget": { "node_count": 2, "item_count": 2, @@ -29380,25 +29282,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs both address chat CLI management URL handling around a `/v1` base, but they do so in different ways and at different scopes: one normalizes the base URL in `RichInterface`, the other rewrites the `load_model` request to target the service root. They are related, but not the same concrete fix.", - "confidence": 0.83, + "summary": "Both PRs address the chat CLI\u2019s management URL construction in `src/transformers/cli/chat.py`, specifically ensuring management endpoints are called against the service root instead of a `/v1` model API base. PR 47303 is the broader, cleaner fix and is the better canonical representative.", + "confidence": 0.87, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #47140 is the cleaner canonical representative: it directly fixes the `/v1` base URL handling at `RichInterface` construction, which is the broader root cause for management URL generation in the chat CLI.", + "canonical_pr_reason": "PR 47303 is the broader and more direct fix: it centralizes service-root derivation for chat management calls and routes the management request there, matching the underlying URL-base bug more generally than the narrower `/v1` normalization in PR 47140.", "best_issue_reason": null, - "best_pr_reason": "PR #47140 is the best overall fit because it addresses the URL normalization problem at the source and is broader than the more localized request-routing change in #47303.", + "best_pr_reason": "PR 47303 is the best standalone representative because it captures the core chat management routing problem in a more general way and is less tied to a single `/v1` edge case.", "soft_edge_verdicts": [ { "left": "pull_request:47140", "right": "pull_request:47303", - "accept": false, - "reason": "Related but not duplicate: #47140 normalizes the chat CLI base URL in the constructor, while #47303 changes only the `load_model` request to use the service root. They touch the same area, but not the same concrete code-path fix." + "accept": true, + "reason": "Same underlying chat CLI management URL bug in the same file: both adjust how management calls are built so they hit the service root rather than the model `/v1` base. The second PR is a broader formulation of the same fix." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: both PRs touch `src/transformers/cli/chat.py` and `tests/cli/test_chat.py`, and both address chat management URL construction around the `/v1` base/service-root split. The soft-edge verdict is conservative enough given the shared files and closely related diffs, and PR 47303 is reasonably described as the broader representative. No obvious overstatement beyond the evidence shown." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -29421,7 +29326,7 @@ "soft_pairs": [ "pull_request:47261|pull_request:47262" ], - "prepared_review_unit_hash": "0c598683f44ce4b25b5bad3e72be71418df5ea1e51108726031fe33fada4869e", + "prepared_review_unit_hash": "35e32d12c736e43f5cb4554ded5978eb283105e173ad90557aa70034e7c5fa58", "component_budget": { "node_count": 2, "item_count": 2, @@ -29443,18 +29348,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two pull requests are unrelated and should not be treated as duplicates: one fixes label smoothing/gradient accumulation loss scaling in training, the other adds/validates min_tokens_to_keep for TopHLogitsWarper in generation.", - "confidence": 0.97, + "summary": "The two pull requests target unrelated bugs in different parts of the library: training loss scaling vs. generation logits warping. They should not be deduplicated or merged together.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47261 is the stronger representative because it addresses a concrete training bug, has an explicit issue target, and the diff clearly implements a focused fix in the loss computation path.", + "canonical_pr_reason": "No single canonical PR for deduplication: PR 47261 fixes gradient inflation with label smoothing during gradient accumulation, while PR 47262 adds/validates `min_tokens_to_keep` in `TopHLogitsWarper`.", "best_issue_reason": null, - "best_pr_reason": "PR 47261 is the best overall representative of the cluster because it is the more complete, issue-linked, and bug-specific change; PR 47262 is a separate generation-side enhancement/fix.", + "best_pr_reason": "PR 47261 is the stronger representative of its own bug report because it has an explicit issue link and directly addresses a concrete training-correctness defect; PR 47262 is a separate generation feature/bugfix.", "soft_edge_verdicts": [ { "left": "pull_request:47261", "right": "pull_request:47262", "accept": false, - "reason": "Different code paths and problems: 47261 changes trainer label smoothing loss handling for gradient accumulation, while 47262 changes generation logits warping and min_tokens_to_keep validation. They do not look mergeable into one PR." + "reason": "Different underlying problems in different code paths (`trainer.py`/`trainer_pt_utils.py` vs `generation/logits_process.py`); they are not the same fix and would not plausibly be merged into one PR." } ] }, @@ -29484,12 +29389,12 @@ "soft_pairs": [ "pull_request:47315|pull_request:47430" ], - "prepared_review_unit_hash": "a30fbbe165a7c8dc0b520795f423703f0173e988f29a8e532aeef223eac16a12", + "prepared_review_unit_hash": "fcb53bfc459fcb5e4e639c2000edb593e5d0969ea6335de551f41e0697c3ebb8", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4383, + "serialized_chars": 4382, "estimated_input_tokens": 1096, "estimated_eval_tokens": 2448 }, @@ -29497,7 +29402,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4383, + "serialized_chars": 4382, "estimated_input_tokens": 1096, "estimated_eval_tokens": 2448 }, @@ -29506,18 +29411,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Two open PRs touch the same compressed_tensors MoE decompression code path, but they fix different bugs: one corrects shape reconstruction for non-32-divisible bit widths, the other preserves zero_point for asymmetric quantization. They are related, not duplicates.", - "confidence": 0.87, + "summary": "These two PRs touch the same compressed_tensors area but fix different bugs: one corrects MoE expert shape reconstruction for non-32-divisible bit widths, the other collects zero_point for asymmetric decompression. They should not be clustered as duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47315 is the clearest representative of the cluster because it addresses a concrete decompression bug in compressed_tensors with a tight, focused fix and explanation.", + "canonical_pr_reason": "No single canonical PR: PR 47315 fixes shape reconstruction for non-32-divisible packing, while PR 47430 fixes missing zero_point handling for asymmetric MoE decompression.", "best_issue_reason": null, - "best_pr_reason": "PR 47315 is the strongest standalone candidate: it fixes a specific code-path failure for expert weight unpacking on 3/5/6/7-bit schemes, and the diff directly targets that behavior.", + "best_pr_reason": "If one PR must represent the cluster, PR 47315 is the better anchor because it addresses a concrete decompression failure mode with a narrowly scoped fix; however, it is not a duplicate of PR 47430.", "soft_edge_verdicts": [ { "left": "pull_request:47315", "right": "pull_request:47430", "accept": false, - "reason": "Different underlying bugs. 47315 fixes incorrect weight_shape fallback for non-32-divisible bit widths; 47430 adds zero_point handling for asymmetric quantization. Same subsystem, but not the same change or mergeable duplicate." + "reason": "Different underlying bugs and fixes: shape fallback for low-bit packing vs. zero_point collection for asymmetric quantization. They share a file and subsystem, but not the same code-path problem." } ] }, @@ -29547,7 +29452,7 @@ "soft_pairs": [ "pull_request:47400|pull_request:47421" ], - "prepared_review_unit_hash": "aa3d72c0e6f19e4ed4af3ab376ac1c94d555c45c5ef6f8d865cbc99be3d0e243", + "prepared_review_unit_hash": "b8517d6f5b6e0ff5cfa2df92dffed83d4325863b9e63b2cc5d43d5ab89afef8c", "component_budget": { "node_count": 2, "item_count": 2, @@ -29569,18 +29474,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate GGUF model-support PRs: one for Phi-2 and one for DeepSeek-V3. They share the same integration/docs/test files, but they target different model-specific code paths and are not duplicates.", - "confidence": 0.95, + "summary": "These are two separate GGUF model-support PRs touching the same integration code, but they add support for different architectures (Phi-2 vs DeepSeek-V3), so they are not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 47421 is the better cluster representative because it is the newer, more substantial GGUF support addition; PR 47400 is a separate Phi-2 support change.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 47421 is the strongest standalone PR here: it adds DeepSeek-V3 GGUF loading support with its own arch-specific changes, docs, and tests.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:47400", "right": "pull_request:47421", "accept": false, - "reason": "Different model targets and code paths (Phi-2 vs DeepSeek-V3). Shared GGUF files reflect common integration points, not the same underlying fix, so they should not be merged as duplicates." + "reason": "Same GGUF subsystem and overlapping files, but the concrete changes target different model architectures and different support paths, so they are not the same underlying change." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index f02db2ff14ffb90904d2f5968d99e2bd024bc868..53f9383d5cf4cd1ca78ff17e7905eb8b33efeff5 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260721t180022z", + "analysis_id": "hybrid-model-20260722t000038z", "archived_artifacts": { - "hybrid": "snapshots/20260721T180022Z/analysis-runs/hybrid-model-20260721t180022z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260721T180022Z/analysis-runs/hybrid-model-20260721t180022z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-07-21T18:20:57Z", + "published_at": "2026-07-22T00:15:46Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260721T180022Z", + "snapshot_id": "20260722T000038Z", "variant": "hybrid" } diff --git a/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..38f068793618d995c2c8fadd750ff597d1562b4c --- /dev/null +++ b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json @@ -0,0 +1,4994 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260722T000038Z", + "generated_at": "2026-07-22T00:15:44Z", + "evidence_quality": "full", + "llm_enrichment": true, + "meta_bugs": [ + { + "cluster_id": "cluster-47122-13", + "summary": "Cluster of 13 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47122, + "issue_numbers": [], + "pr_numbers": [ + 47122, + 47123, + 47124, + 47224, + 47266, + 47280, + 47281, + 47282, + 47310, + 47380, + 47443, + 47444, + 47445 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47122, + "right_pr_number": 47123, + "code_similarity": 0.092, + "size_similarity": 0.462, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47124, + "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": 47122, + "right_pr_number": 47224, + "code_similarity": 0.474, + "size_similarity": 0.302, + "file_overlap": 0.4, + "area_overlap": 0.818, + "patch_similarity": 0.051, + "shared_filenames": [ + "src/transformers/models/edgetam/modeling_edgetam.py", + "src/transformers/models/edgetam/modular_edgetam.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/edgetam/modeling_edgetam.py", + "left_ranges": [ + [ + 464, + 472 + ] + ], + "right_ranges": [ + [ + 464, + 474 + ] + ] + }, + { + "filename": "src/transformers/models/edgetam/modular_edgetam.py", + "left_ranges": [ + [ + 210, + 218 + ] + ], + "right_ranges": [ + [ + 210, + 220 + ] + ] + } + ] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47266, + "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": 47122, + "right_pr_number": 47280, + "code_similarity": 0.177, + "size_similarity": 0.885, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47281, + "code_similarity": 0.05, + "size_similarity": 0.228, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.03, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47282, + "code_similarity": 0.118, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47310, + "code_similarity": 0.111, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.074, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47380, + "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": 47122, + "right_pr_number": 47443, + "code_similarity": 0.011, + "size_similarity": 0.051, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.006, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47444, + "code_similarity": 0.02, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.023, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47122, + "right_pr_number": 47445, + "code_similarity": 0.02, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.025, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47124, + "code_similarity": 0.092, + "size_similarity": 0.462, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47224, + "code_similarity": 0.028, + "size_similarity": 0.14, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47266, + "code_similarity": 0.159, + "size_similarity": 0.545, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.333, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47280, + "code_similarity": 0.547, + "size_similarity": 0.522, + "file_overlap": 1.0, + "area_overlap": 0.368, + "patch_similarity": 0.093, + "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, + 21 + ], + [ + 481, + 499 + ], + [ + 520, + 530 + ] + ] + } + ] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47281, + "code_similarity": 0.023, + "size_similarity": 0.105, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.011, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47282, + "code_similarity": 0.055, + "size_similarity": 0.273, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47310, + "code_similarity": 0.185, + "size_similarity": 0.923, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47380, + "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": 47123, + "right_pr_number": 47443, + "code_similarity": 0.006, + "size_similarity": 0.023, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.01, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47444, + "code_similarity": 0.009, + "size_similarity": 0.037, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.013, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47123, + "right_pr_number": 47445, + "code_similarity": 0.01, + "size_similarity": 0.037, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.014, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47224, + "code_similarity": 0.06, + "size_similarity": 0.302, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47266, + "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": 47124, + "right_pr_number": 47280, + "code_similarity": 0.179, + "size_similarity": 0.885, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.013, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47281, + "code_similarity": 0.048, + "size_similarity": 0.228, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.016, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47282, + "code_similarity": 0.128, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.067, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47310, + "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": 47124, + "right_pr_number": 47380, + "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": 47124, + "right_pr_number": 47443, + "code_similarity": 0.014, + "size_similarity": 0.051, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.025, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47444, + "code_similarity": 0.016, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47124, + "right_pr_number": 47445, + "code_similarity": 0.016, + "size_similarity": 0.08, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47266, + "code_similarity": 0.056, + "size_similarity": 0.256, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47280, + "code_similarity": 0.058, + "size_similarity": 0.267, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47281, + "code_similarity": 0.159, + "size_similarity": 0.754, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.052, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47282, + "code_similarity": 0.105, + "size_similarity": 0.512, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.018, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47310, + "code_similarity": 0.037, + "size_similarity": 0.151, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.047, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47380, + "code_similarity": 0.056, + "size_similarity": 0.256, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.032, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47443, + "code_similarity": 0.039, + "size_similarity": 0.168, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47444, + "code_similarity": 0.058, + "size_similarity": 0.266, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.033, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47224, + "right_pr_number": 47445, + "code_similarity": 0.058, + "size_similarity": 0.266, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47280, + "code_similarity": 0.229, + "size_similarity": 0.957, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.25, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47281, + "code_similarity": 0.045, + "size_similarity": 0.193, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.04, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47282, + "code_similarity": 0.104, + "size_similarity": 0.5, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.028, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47310, + "code_similarity": 0.118, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.0, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47380, + "code_similarity": 0.226, + "size_similarity": 1.0, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.174, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47443, + "code_similarity": 0.177, + "size_similarity": 0.043, + "file_overlap": 0.1, + "area_overlap": 0.387, + "patch_similarity": 0.02, + "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, + 20 + ], + [ + 245, + 264 + ], + [ + 277, + 283 + ], + [ + 298, + 304 + ], + [ + 339, + 345 + ] + ], + "right_ranges": [ + [ + 13, + 19 + ], + [ + 245, + 360 + ] + ] + } + ] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47444, + "code_similarity": 0.019, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47266, + "right_pr_number": 47445, + "code_similarity": 0.021, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.052, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47281, + "code_similarity": 0.046, + "size_similarity": 0.202, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47282, + "code_similarity": 0.107, + "size_similarity": 0.523, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.02, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47310, + "code_similarity": 0.116, + "size_similarity": 0.565, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.017, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47380, + "code_similarity": 0.205, + "size_similarity": 0.957, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.094, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47443, + "code_similarity": 0.014, + "size_similarity": 0.045, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.036, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47444, + "code_similarity": 0.018, + "size_similarity": 0.071, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47280, + "right_pr_number": 47445, + "code_similarity": 0.02, + "size_similarity": 0.071, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.037, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47282, + "code_similarity": 0.082, + "size_similarity": 0.386, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.034, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47310, + "code_similarity": 0.026, + "size_similarity": 0.114, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.019, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47380, + "code_similarity": 0.046, + "size_similarity": 0.193, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.049, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47443, + "code_similarity": 0.052, + "size_similarity": 0.223, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.051, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47444, + "code_similarity": 0.082, + "size_similarity": 0.353, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.076, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47281, + "right_pr_number": 47445, + "code_similarity": 0.084, + "size_similarity": 0.353, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.088, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47310, + "code_similarity": 0.063, + "size_similarity": 0.295, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.026, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47380, + "code_similarity": 0.476, + "size_similarity": 0.5, + "file_overlap": 1.0, + "area_overlap": 0.194, + "patch_similarity": 0.056, + "shared_filenames": [ + "tests/models/mamba2/test_modeling_mamba2.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/mamba2/test_modeling_mamba2.py", + "left_ranges": [ + [ + 420, + 438 + ], + [ + 450, + 474 + ], + [ + 482, + 524 + ] + ], + "right_ranges": [ + [ + 18, + 24 + ], + [ + 406, + 419 + ], + [ + 426, + 432 + ], + [ + 457, + 463 + ], + [ + 487, + 493 + ] + ] + } + ] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47443, + "code_similarity": 0.027, + "size_similarity": 0.086, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.068, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47444, + "code_similarity": 0.032, + "size_similarity": 0.136, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.03, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47282, + "right_pr_number": 47445, + "code_similarity": 0.033, + "size_similarity": 0.136, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.04, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47380, + "code_similarity": 0.123, + "size_similarity": 0.591, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.032, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47443, + "code_similarity": 0.008, + "size_similarity": 0.025, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.023, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47444, + "code_similarity": 0.015, + "size_similarity": 0.04, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.044, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47310, + "right_pr_number": 47445, + "code_similarity": 0.015, + "size_similarity": 0.04, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.049, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47443, + "code_similarity": 0.013, + "size_similarity": 0.043, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.03, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47444, + "code_similarity": 0.02, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.045, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47380, + "right_pr_number": 47445, + "code_similarity": 0.023, + "size_similarity": 0.068, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.063, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47443, + "right_pr_number": 47444, + "code_similarity": 0.754, + "size_similarity": 0.632, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.217, + "shared_filenames": [ + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "left_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ], + "right_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ] + }, + { + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "left_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ], + "right_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ], + "right_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ], + "right_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ], + "right_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ] + }, + { + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "left_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ], + "right_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ], + "right_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ] + }, + { + "filename": "src/transformers/models/youtu/modeling_youtu.py", + "left_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ], + "right_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ] + } + ] + }, + { + "left_pr_number": 47443, + "right_pr_number": 47445, + "code_similarity": 0.755, + "size_similarity": 0.632, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.219, + "shared_filenames": [ + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "left_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ], + "right_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ] + }, + { + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "left_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ], + "right_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ], + "right_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ], + "right_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ], + "right_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ] + }, + { + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "left_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ], + "right_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ], + "right_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ] + }, + { + "filename": "src/transformers/models/youtu/modeling_youtu.py", + "left_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ], + "right_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ] + } + ] + }, + { + "left_pr_number": 47444, + "right_pr_number": 47445, + "code_similarity": 0.916, + "size_similarity": 1.0, + "file_overlap": 0.818, + "area_overlap": 1.0, + "patch_similarity": 0.802, + "shared_filenames": [ + "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/mistral4/modeling_mistral4.py", + "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "src/transformers/models/youtu/modeling_youtu.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", + "left_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ], + "right_ranges": [ + [ + 4, + 9 + ], + [ + 310, + 315 + ] + ] + }, + { + "filename": "src/transformers/models/deepseek_v32/modeling_deepseek_v32.py", + "left_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ], + "right_ranges": [ + [ + 299, + 304 + ], + [ + 338, + 349 + ], + [ + 392, + 403 + ] + ] + }, + { + "filename": "src/transformers/models/glm4_moe_lite/modeling_glm4_moe_lite.py", + "left_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ], + "right_ranges": [ + [ + 182, + 187 + ], + [ + 221, + 232 + ], + [ + 272, + 283 + ] + ] + }, + { + "filename": "src/transformers/models/glm_moe_dsa/modeling_glm_moe_dsa.py", + "left_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ], + "right_ranges": [ + [ + 306, + 311 + ], + [ + 358, + 369 + ] + ] + }, + { + "filename": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", + "left_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ], + "right_ranges": [ + [ + 282, + 287 + ], + [ + 321, + 332 + ], + [ + 372, + 383 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ], + "right_ranges": [ + [ + 196, + 201 + ], + [ + 283, + 288 + ] + ] + }, + { + "filename": "src/transformers/models/mistral4/modeling_mistral4.py", + "left_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ], + "right_ranges": [ + [ + 17, + 22 + ], + [ + 319, + 324 + ] + ] + }, + { + "filename": "src/transformers/models/nemotron_h/modeling_nemotron_h.py", + "left_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ], + "right_ranges": [ + [ + 109, + 114 + ], + [ + 291, + 300 + ], + [ + 361, + 370 + ], + [ + 396, + 409 + ], + [ + 435, + 444 + ] + ] + }, + { + "filename": "src/transformers/models/youtu/modeling_youtu.py", + "left_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ], + "right_ranges": [ + [ + 222, + 227 + ], + [ + 261, + 272 + ], + [ + 312, + 323 + ] + ] + } + ] + } + ] + }, + { + "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-43451-5", + "summary": "Cluster of 5 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, + 45157, + 46509, + 47163, + 47306 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43451, + "right_pr_number": 45157, + "code_similarity": 0.016, + "size_similarity": 0.036, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.058, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43451, + "right_pr_number": 46509, + "code_similarity": 0.138, + "size_similarity": 0.363, + "file_overlap": 0.097, + "area_overlap": 0.0, + "patch_similarity": 0.243, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 1381, + 1388 + ] + ], + "right_ranges": [ + [ + 795, + 802 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 398, + 405 + ], + [ + 841, + 847 + ], + [ + 965, + 971 + ], + [ + 1089, + 1095 + ], + [ + 1225, + 1231 + ] + ], + "right_ranges": [ + [ + 412, + 418 + ], + [ + 1077, + 1083 + ], + [ + 1206, + 1212 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 335, + 342 + ], + [ + 1107, + 1113 + ], + [ + 1930, + 1936 + ] + ], + "right_ranges": [ + [ + 1097, + 1103 + ] + ] + } + ] + }, + { + "left_pr_number": 43451, + "right_pr_number": 47163, + "code_similarity": 0.004, + "size_similarity": 0.004, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.02, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43451, + "right_pr_number": 47306, + "code_similarity": 0.04, + "size_similarity": 0.122, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.107, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 46509, + "code_similarity": 0.032, + "size_similarity": 0.101, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.08, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 47163, + "code_similarity": 0.029, + "size_similarity": 0.117, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.035, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 45157, + "right_pr_number": 47306, + "code_similarity": 0.078, + "size_similarity": 0.298, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.125, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47163, + "code_similarity": 0.006, + "size_similarity": 0.012, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.024, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 46509, + "right_pr_number": 47306, + "code_similarity": 0.086, + "size_similarity": 0.337, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.126, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47163, + "right_pr_number": 47306, + "code_similarity": 0.011, + "size_similarity": 0.035, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-41211-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41211.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41211, + "canonical_pr_number": 44339, + "issue_numbers": [ + 41211 + ], + "pr_numbers": [ + 41356, + 44339 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41356, + "right_pr_number": 44339, + "code_similarity": 0.155, + "size_similarity": 0.094, + "file_overlap": 0.273, + "area_overlap": 0.078, + "patch_similarity": 0.183, + "shared_filenames": [ + "docs/source/en/model_doc/deimv2.md", + "src/transformers/models/deimv2/__init__.py", + "src/transformers/models/deimv2/configuration_deimv2.py", + "src/transformers/models/deimv2/modeling_deimv2.py", + "tests/models/deimv2/__init__.py", + "tests/models/deimv2/test_modeling_deimv2.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/model_doc/deimv2.md", + "left_ranges": [ + [ + 1, + 132 + ] + ], + "right_ranges": [ + [ + 1, + 65 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/__init__.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/configuration_deimv2.py", + "left_ranges": [ + [ + 1, + 74 + ] + ], + "right_ranges": [ + [ + 1, + 266 + ] + ] + }, + { + "filename": "tests/models/deimv2/test_modeling_deimv2.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 1734 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46894-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #46894.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 46894, + "canonical_pr_number": 47090, + "issue_numbers": [ + 46894, + 46897 + ], + "pr_numbers": [ + 46958, + 47090 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46958, + "right_pr_number": 47090, + "code_similarity": 0.397, + "size_similarity": 0.257, + "file_overlap": 0.273, + "area_overlap": 0.598, + "patch_similarity": 0.362, + "shared_filenames": [ + "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "tests/models/cohere_asr/test_modeling_cohere_asr.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/cohere_asr/modeling_cohere_asr.py", + "left_ranges": [ + [ + 22, + 28 + ], + [ + 638, + 645 + ] + ], + "right_ranges": [ + [ + 637, + 649 + ] + ] + }, + { + "filename": "src/transformers/models/cohere_asr/modular_cohere_asr.py", + "left_ranges": [ + [ + 16, + 22 + ], + [ + 501, + 508 + ] + ], + "right_ranges": [ + [ + 500, + 512 + ] + ] + }, + { + "filename": "tests/models/cohere_asr/test_modeling_cohere_asr.py", + "left_ranges": [ + [ + 15, + 21 + ], + [ + 180, + 221 + ] + ], + "right_ranges": [ + [ + 175, + 224 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-47301-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #47333.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 47333, + "canonical_pr_number": 47336, + "issue_numbers": [ + 47332, + 47333 + ], + "pr_numbers": [ + 47301, + 47336 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47301, + "right_pr_number": 47336, + "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": [] + } + ] + }, + { + "cluster_id": "cluster-47138-4", + "summary": "Cluster of 2 issues and 2 PRs centered on issue #47138.", + "status": "open", + "confidence": 0.6, + "canonical_issue_number": 47138, + "canonical_pr_number": 47303, + "issue_numbers": [ + 47138, + 47139 + ], + "pr_numbers": [ + 47140, + 47303 + ], + "evidence_types": [ + "closing_reference", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47140, + "right_pr_number": 47303, + "code_similarity": 0.656, + "size_similarity": 0.65, + "file_overlap": 1.0, + "area_overlap": 0.447, + "patch_similarity": 0.466, + "shared_filenames": [ + "src/transformers/cli/chat.py", + "tests/cli/test_chat.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cli/chat.py", + "left_ranges": [ + [ + 20, + 26 + ], + [ + 103, + 122 + ], + [ + 346, + 356 + ], + [ + 383, + 389 + ], + [ + 473, + 479 + ] + ], + "right_ranges": [ + [ + 103, + 115 + ], + [ + 216, + 224 + ], + [ + 383, + 389 + ] + ] + }, + { + "filename": "tests/cli/test_chat.py", + "left_ranges": [ + [ + 15, + 28 + ], + [ + 51, + 105 + ] + ], + "right_ranges": [ + [ + 14, + 29 + ], + [ + 52, + 76 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46658-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46658, + "issue_numbers": [], + "pr_numbers": [ + 46658, + 46836 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46658, + "right_pr_number": 46836, + "code_similarity": 0.206, + "size_similarity": 0.71, + "file_overlap": 0.059, + "area_overlap": 0.0, + "patch_similarity": 0.308, + "shared_filenames": [ + "src/transformers/conversion_mapping.py", + "src/transformers/models/auto/auto_mappings.py", + "src/transformers/models/auto/modeling_auto.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/conversion_mapping.py", + "left_ranges": [ + [ + 739, + 830 + ] + ], + "right_ranges": [ + [ + 127, + 133 + ], + [ + 1495, + 1545 + ] + ] + }, + { + "filename": "src/transformers/models/auto/auto_mappings.py", + "left_ranges": [ + [ + 362, + 368 + ], + [ + 591, + 599 + ], + [ + 798, + 804 + ], + [ + 884, + 891 + ], + [ + 1084, + 1090 + ], + [ + 1207, + 1213 + ] + ], + "right_ranges": [ + [ + 637, + 647 + ], + [ + 919, + 928 + ], + [ + 1130, + 1136 + ], + [ + 1260, + 1266 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 1085, + 1091 + ] + ], + "right_ranges": [ + [ + 509, + 515 + ], + [ + 1119, + 1125 + ] + ] + } + ] + } + ] + }, + { + "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-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-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-45561-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45561.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 45561, + "canonical_pr_number": 45645, + "issue_numbers": [ + 45561 + ], + "pr_numbers": [ + 45639, + 45645 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45639, + "right_pr_number": 45645, + "code_similarity": 0.721, + "size_similarity": 0.973, + "file_overlap": 0.75, + "area_overlap": 0.563, + "patch_similarity": 0.696, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py", + "utils/notification_service.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3611, + 3618 + ], + [ + 3835, + 3869 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 144 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + }, + { + "filename": "utils/notification_service.py", + "left_ranges": [ + [ + 935, + 967 + ] + ], + "right_ranges": [ + [ + 935, + 967 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46614-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46614.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46614, + "canonical_pr_number": 46687, + "issue_numbers": [ + 46614 + ], + "pr_numbers": [ + 46687, + 47461 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46687, + "right_pr_number": 47461, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/models/wav2vec2_phoneme/tokenization_wav2vec2_phoneme.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/wav2vec2_phoneme/tokenization_wav2vec2_phoneme.py", + "left_ranges": [ + [ + 223, + 229 + ], + [ + 255, + 261 + ] + ], + "right_ranges": [ + [ + 223, + 229 + ], + [ + 255, + 261 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46305-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46305.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 46305, + "canonical_pr_number": 47162, + "issue_numbers": [ + 46305 + ], + "pr_numbers": [ + 46310, + 47162 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46310, + "right_pr_number": 47162, + "code_similarity": 0.577, + "size_similarity": 0.338, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.062, + "shared_filenames": [ + "src/transformers/models/moonshine_streaming/modeling_moonshine_streaming.py", + "src/transformers/models/moonshine_streaming/modular_moonshine_streaming.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/moonshine_streaming/modeling_moonshine_streaming.py", + "left_ranges": [ + [ + 356, + 362 + ] + ], + "right_ranges": [ + [ + 356, + 362 + ] + ] + }, + { + "filename": "src/transformers/models/moonshine_streaming/modular_moonshine_streaming.py", + "left_ranges": [ + [ + 269, + 275 + ] + ], + "right_ranges": [ + [ + 269, + 275 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-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-47173-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47449, + "issue_numbers": [], + "pr_numbers": [ + 47173, + 47292, + 47449 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 47173, + "right_pr_number": 47292, + "code_similarity": 0.097, + "size_similarity": 0.463, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.027, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47173, + "right_pr_number": 47449, + "code_similarity": 0.1, + "size_similarity": 0.47, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.038, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 47292, + "right_pr_number": 47449, + "code_similarity": 0.201, + "size_similarity": 0.984, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.029, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-46912-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 47175, + "issue_numbers": [], + "pr_numbers": [ + 46912, + 47175 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 46912, + "right_pr_number": 47175, + "code_similarity": 0.009, + "size_similarity": 0.005, + "file_overlap": 0.022, + "area_overlap": 0.0, + "patch_similarity": 0.01, + "shared_filenames": [ + "tests/models/olmo3/test_modeling_olmo3.py" + ], + "shared_file_areas": [ + { + "filename": "tests/models/olmo3/test_modeling_olmo3.py", + "left_ranges": [ + [ + 99, + 104 + ] + ], + "right_ranges": [ + [ + 213, + 227 + ], + [ + 312, + 318 + ] + ] + } + ] + } + ] + }, + { + "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, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47138-4", + "canonical_issue_number": 47138, + "duplicate_issue_numbers": [ + 47139 + ], + "reason": "Issues in cluster-47138-4 are treated as duplicates because they share closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47301-4", + "canonical_issue_number": 47333, + "duplicate_issue_numbers": [ + 47332 + ], + "reason": "Issues in cluster-47301-4 are treated as duplicates because they share closing_reference, 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-43451-5", + "canonical_pr_number": 43451, + "duplicate_pr_numbers": [ + 45157, + 46509, + 47163, + 47306 + ], + "target_issue_number": null, + "reason": "PRs in cluster-43451-5 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, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45647-3", + "canonical_pr_number": 45738, + "duplicate_pr_numbers": [ + 45737 + ], + "target_issue_number": 45647, + "reason": "PRs in cluster-45647-3 are treated as duplicates because they converge on issue #45647 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46159-3", + "canonical_pr_number": 46174, + "duplicate_pr_numbers": [ + 46780 + ], + "target_issue_number": 46159, + "reason": "PRs in cluster-46159-3 are treated as duplicates because they converge on issue #46159 with closing_reference evidence." + }, + { + "cluster_id": "cluster-46305-3", + "canonical_pr_number": 47162, + "duplicate_pr_numbers": [ + 46310 + ], + "target_issue_number": 46305, + "reason": "PRs in cluster-46305-3 are treated as duplicates because they converge on issue #46305 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46614-3", + "canonical_pr_number": 46687, + "duplicate_pr_numbers": [ + 47461 + ], + "target_issue_number": 46614, + "reason": "PRs in cluster-46614-3 are treated as duplicates because they converge on issue #46614 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46658-2", + "canonical_pr_number": 46658, + "duplicate_pr_numbers": [ + 46836 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46658-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-46869-3", + "canonical_pr_number": 46886, + "duplicate_pr_numbers": [ + 46922 + ], + "target_issue_number": 46869, + "reason": "PRs in cluster-46869-3 are treated as duplicates because they converge on issue #46869 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46894-4", + "canonical_pr_number": 47090, + "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, soft_similarity evidence." + }, + { + "cluster_id": "cluster-46912-2", + "canonical_pr_number": 47175, + "duplicate_pr_numbers": [ + 46912 + ], + "target_issue_number": null, + "reason": "PRs in cluster-46912-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47122-13", + "canonical_pr_number": 47122, + "duplicate_pr_numbers": [ + 47123, + 47124, + 47224, + 47266, + 47280, + 47281, + 47282, + 47310, + 47380, + 47443, + 47444, + 47445 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47122-13 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47138-4", + "canonical_pr_number": 47303, + "duplicate_pr_numbers": [ + 47140 + ], + "target_issue_number": 47138, + "reason": "PRs in cluster-47138-4 are treated as duplicates because they converge on issue #47138 with closing_reference, soft_similarity evidence." + }, + { + "cluster_id": "cluster-47173-3", + "canonical_pr_number": 47449, + "duplicate_pr_numbers": [ + 47173, + 47292 + ], + "target_issue_number": null, + "reason": "PRs in cluster-47173-3 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-47301-4", + "canonical_pr_number": 47336, + "duplicate_pr_numbers": [ + 47301 + ], + "target_issue_number": 47333, + "reason": "PRs in cluster-47301-4 are treated as duplicates because they converge on issue #47333 with closing_reference, soft_similarity evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-47301-4", + "issue_number": 47333, + "reason": "Issue #47333 is the strongest global issue candidate because it is open, belongs to a cluster with 4 artifacts, and carries 5 discussion comments plus 4 inbound references.", + "score": 154.0 + }, + "best_pr": { + "cluster_id": "cluster-47122-13", + "pr_number": 47122, + "reason": "PR #47122 is the strongest global PR candidate because it is open, belongs to a cluster with 13 artifacts, links to 0 issues, and carries 8 review/discussion events.", + "score": 265.974 + } +} diff --git a/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..b38a8f157556a1d2ad38dd83c26d6a82c4f95b2a --- /dev/null +++ b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,29500 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260722T000038Z", + "generated_at": "2026-07-22T00:15:44Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 1, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34928", + "issue:35308", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:39921", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43425", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "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:39921|issue:44918", + "issue:43828|issue:45237", + "issue:39921|issue:43039", + "issue:45237|issue:46468", + "issue:39921|issue:45685", + "issue:45237|issue:45507", + "issue:35545|issue:39921", + "issue:30277|issue:35545", + "issue:39921|issue:44805", + "issue:39921|issue:43716", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:35545|issue:36010", + "issue:44918|issue:46165", + "issue:44509|issue:44512", + "issue:46165|issue:46869", + "issue:44805|issue:45823", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:46294", + "issue:46165|issue:46468", + "issue:44805|issue:45161", + "issue:44560|issue:46165", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43039|issue:45161", + "issue:43975|issue:44457", + "issue:44292|issue:44457", + "issue:34928|issue:35308", + "issue:44841|issue:46661", + "issue:24643|issue:30277", + "issue:43716|issue:46869", + "issue:34928|issue:44805", + "issue:45685|issue:46869", + "issue:34928|issue:36331", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:39921", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:44292|issue:45237", + "issue:45874|issue:46294", + "issue:39290|issue:45081", + "issue:45800|issue:46515", + "issue:34928|issue:41720", + "issue:45823|issue:45874", + "issue:43425|issue:44292", + "issue:43828|issue:45161", + "issue:43541|issue:43716", + "issue:43828|issue:44292", + "issue:43716|issue:45685", + "issue:45507|issue:45685", + "issue:44292|issue:44918", + "issue:43716|issue:44292", + "issue:44918|issue:45161", + "issue:43039|issue:44918", + "issue:41762|issue:43638" + ], + "prepared_review_unit_hash": "0bb96df5b444835f6f3bd8214c8d2c1027165b24e2cae5705a30365804380bc5", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 62, + "serialized_chars": 46364, + "estimated_input_tokens": 11591, + "estimated_eval_tokens": 23438 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. The only clear duplicate subclusters are the Zero-3/BERT index-error reports, the v5 docs cleanup reports, and a couple of closely related torch/version or Qwen3.5-MoE embedding issues.", + "confidence": 0.78, + "canonical_issue_reason": "issue:43596 is the clearest representative of the strongest duplicate pair (with 43638) and has the most concrete, reproducible Zero-3/BERT crash signature.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43596 is the most actionable standalone bug report in the set: precise error, clear code path, and a near-identical duplicate companion.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different failure modes and code paths: config attribute error vs streaming chat/completions inference." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different subsystems and bugs: AMD GPU failure vs MPS histogram op limitation." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different models and root causes: image-preprocessor dtype mismatch vs AMD GPU support issue." + }, + { + "left": "issue:39921", + "right": "issue:44918", + "accept": false, + "reason": "Different models and errors: special-token projection issue vs TRL SFT embedding unpacking." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Both are dtype/GPU-related, but they describe different models and concrete failures." + }, + { + "left": "issue:39921", + "right": "issue:43039", + "accept": false, + "reason": "Gemma3N token projection bug is unrelated to Liger Kernel cross-entropy dispatch." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Unrelated model-specific bugs with no shared concrete code path." + }, + { + "left": "issue:39921", + "right": "issue:45685", + "accept": false, + "reason": "Different model families and different failures; no shared underlying change." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different model/runtime failures; both MoE-ish at a high level but not the same bug." + }, + { + "left": "issue:35545", + "right": "issue:39921", + "accept": false, + "reason": "Unrelated issues: ONNX export vs special-token projection handling." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collective mismatch and ONNX export error are different code paths." + }, + { + "left": "issue:39921", + "right": "issue:44805", + "accept": false, + "reason": "Different bug classes: token projection vs tensor mask shape mismatch." + }, + { + "left": "issue:39921", + "right": "issue:43716", + "accept": false, + "reason": "Different model-specific failures with no shared code path." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "ONNX export and FineGrainedFP8 model execution are unrelated." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "Different models and failures; no concrete overlap beyond general inference tooling." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "One is a table-QA crash, the other is outdated RAG documentation." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "ONNX export bug is unrelated to the GenerationMixin import regression." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "Different training/resume issues with no shared concrete bug." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": true, + "reason": "Both are v5 docs cleanup reports about stale removed commands/tasks; same documentation-maintenance change." + }, + { + "left": "issue:46165", + "right": "issue:46869", + "accept": false, + "reason": "Different problems: resume-with-remote-code vs multi-GPU YOLO device-map failure." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Mask-shape mismatch and Gemma4 device mismatch are distinct runtime errors." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Different error signatures and different model paths." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Zero-3/BERT index error is unrelated to the mask-indexing shape mismatch." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Different modalities and failure sites; not the same underlying bug." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Unrelated issues with no shared code path." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Mask-shape indexing bug vs GPT-OSS MoE TP failure are different problems." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Different feature areas: Qwen3-VL chat-template video error vs training resume issue." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Both involve DeepSpeed, but the concrete failures and models differ." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same underlying bug: DeepSpeed ZeRO-3/BERT loading hits the same index-0-out-of-bounds crash." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Liger Kernel cross-entropy dispatch is unrelated to GPT-OSS MoE tensor-parallel failure." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Different tasks: detokenization regression vs LoRA merge/save/load inconsistency." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Different model/runtime failures; no shared code path." + }, + { + "left": "issue:34928", + "right": "issue:35308", + "accept": false, + "reason": "Both are multi-GPU/training issues, but the symptoms and causes are different." + }, + { + "left": "issue:44841", + "right": "issue:46661", + "accept": false, + "reason": "Processor failure and remote-code model_class/config_class error are unrelated." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "Both mention DeepSpeed, but the actual failures are different and model-specific." + }, + { + "left": "issue:43716", + "right": "issue:46869", + "accept": false, + "reason": "Different models and different hardware/device-map failures." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpointing/FSDP size mismatch is not the same as a mask indexing error." + }, + { + "left": "issue:45685", + "right": "issue:46869", + "accept": false, + "reason": "MPS histogram limitation is unrelated to multi-GPU device-map auto behavior." + }, + { + "left": "issue:34928", + "right": "issue:36331", + "accept": false, + "reason": "Different training bugs: recomputed-tensor mismatch vs custom trainer signature change." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": true, + "reason": "Same Qwen3.5-MoE rotary-embedding area with offload/materialization bugs; likely the same underlying fix." + }, + { + "left": "issue:43638", + "right": "issue:45161", + "accept": false, + "reason": "Zero-3/BERT crash is unrelated to GPT-OSS MoE tensor-parallel failure." + }, + { + "left": "issue:34928", + "right": "issue:39921", + "accept": false, + "reason": "FSDP/activation-checkpointing issue is unrelated to Gemma3N special-token projection." + }, + { + "left": "issue:43596", + "right": "issue:45161", + "accept": false, + "reason": "Different model families and different runtime paths." + }, + { + "left": "issue:43716", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific failures with no shared concrete bug." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Different model/runtime issues; no same code path." + }, + { + "left": "issue:45874", + "right": "issue:46294", + "accept": false, + "reason": "Expensive reverse-embedding lookup and video chat-template token mismatch are unrelated." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Different bugs: Gemma3 text config regression vs Mistral regex patch crash." + }, + { + "left": "issue:45800", + "right": "issue:46515", + "accept": true, + "reason": "Both report the same torch 2.4.1 incompatibility regression; one is just more specific." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "Different model/runtime failures with no shared code path." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Device-map forward mismatch and reverse-embedding lookup are different bugs." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Generic torch incompatibility and Qwen3-8B NVFP4 runtime error are not the same issue." + }, + { + "left": "issue:43828", + "right": "issue:45161", + "accept": false, + "reason": "Different model families and different dtype/device-parallel failures." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "TorchDynamo grouped_mm tracing error is unrelated to the Mistral image dtype mismatch." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Different model/runtime bugs with no shared underlying change." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Different models and backends; not the same bug." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Different code paths and error types; only broad MoE/backend overlap." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Different model-specific runtime failures." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different models and errors; no concrete duplicate signal." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "TRl SFT embedding unpacking and GPT-OSS TP failure are unrelated." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger Kernel cross-entropy issue is unrelated to Qwen3.5 embedding unpacking." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading failure and BERT ZeRO-3 loading failure are similar in shape, but not the same concrete bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet: the cluster is mostly heterogeneous, and the identified duplicate subclusters are the ones with strong title/body similarity (Zero-3/BERT index errors, v5 docs cleanup, torch 2.4.1 incompatibility, and the Qwen3.5-MoE rotary-embedding pair). The soft-edge rejections are generally conservative and tied to distinct titles/error modes. No clear overstatement stands out." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 2, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:30277", + "issue:35308", + "issue:36331", + "issue:39290", + "issue:41093", + "issue:41720", + "issue:41762", + "issue:42617", + "issue:43039", + "issue:43296", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43901", + "issue:43975", + "issue:44292", + "issue:44512", + "issue:44589", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:45020", + "issue:45070", + "issue:45237", + "issue:45800", + "issue:45812", + "issue:45823", + "issue:45901", + "issue:45920", + "issue:46015", + "issue:46294", + "issue:46304", + "issue:46402", + "issue:46772", + "issue:46869", + "issue:46901", + "issue:47228", + "issue:47394", + "issue:47448" + ], + "soft_pairs": [ + "issue:43827|issue:44512", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:45237|issue:46015", + "issue:43039|issue:44805", + "issue:45800|issue:46304", + "issue:43296|issue:46304", + "issue:39290|issue:41720", + "issue:43866|issue:46402", + "issue:44841|issue:47448", + "issue:44589|issue:46304", + "issue:45237|issue:45901", + "issue:41762|issue:43596", + "issue:43901|issue:44512", + "issue:44841|issue:45823", + "issue:45823|issue:46294", + "issue:43975|issue:46015", + "issue:45020|issue:46304", + "issue:44863|issue:46015", + "issue:43866|issue:46772", + "issue:35308|issue:42617", + "issue:45812|issue:45920", + "issue:43872|issue:46304", + "issue:44292|issue:46015", + "issue:46901|issue:47394", + "issue:39290|issue:43531", + "issue:43866|issue:44863", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43425|issue:45070", + "issue:43854|issue:43866", + "issue:43541|issue:47228", + "issue:43792|issue:45070", + "issue:43975|issue:45901", + "issue:43596|issue:45800", + "issue:43541|issue:46869" + ], + "prepared_review_unit_hash": "5fa8fb3c0efd610dd3510ad47a55d951489cafb5a056ed258823d2a97636d60d", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 36, + "serialized_chars": 37776, + "estimated_input_tokens": 9444, + "estimated_eval_tokens": 19144 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly heterogeneous; only two soft duplicate pairs look convincing: a DeepSpeed ZeRO-3 zero-sized-index loading failure and an AutoTokenizer v5 regression producing wrong token IDs. No PRs are present.", + "confidence": 0.58, + "canonical_issue_reason": "issue:45920 is the broadest and most representative standalone report of the tokenizer regression, covering multiple model families rather than a single model-specific symptom.", + "canonical_pr_reason": "No PRs in the cluster.", + "best_issue_reason": "issue:45920 has the clearest cross-model scope and discussion for the tokenizer ID regression, making it the best issue representative among the accepted duplicate subgroup.", + "best_pr_reason": "No PRs available.", + "soft_edge_verdicts": [ + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs cleanup, but they refer to different removed behaviors/commands and do not describe the same concrete change." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Different bugs: Liger cross-entropy routing vs an autocast dtype mismatch in Phi-tiny-MoE." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "Different failure modes: DeepSpeed collective mismatch vs a trainer signature TypeError." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "Unrelated: AMD GPU runtime failure vs outdated RAG docs." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Different code paths: Liger kernel cross-entropy vs a mask/tensor shape mismatch." + }, + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Both mention compatibility, but these are distinct regressions in different environments and code paths." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Unrelated model-loading/vLLM failure vs a USE_DISTRIBUTED=0 build/runtime regression." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Different problems: missing Gemma3 config attribute in vLLM vs Qwen3 auto device-mapping CUDA assert." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "Different model-specific load failures; no shared underlying bug is evident." + }, + { + "left": "issue:44841", + "right": "issue:47448", + "accept": false, + "reason": "Both are model-architecture issues, but they concern different models and different loading/processor failures." + }, + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Different regressions: missing Float8 storage type vs distributed build support breakage." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Unrelated runtime failure vs table-question-answering crash." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": true, + "reason": "Same underlying DeepSpeed ZeRO-3 zero-sized-index failure during model loading; only the affected model differs." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Separate documentation issues referencing different removed APIs/commands." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Different model-specific failures: processor loading vs device-mapping forward mismatch." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Both are multimodal issues, but they target different mechanisms (device mapping vs chat-template video token alignment)." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Different topics: detokenization bug vs outdated RAG examples." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Remote_code model regressions are not the same as the USE_DISTRIBUTED=0 compatibility break." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "Checkpoint-loading issue vs documentation drift; unrelated." + }, + { + "left": "issue:43866", + "right": "issue:46772", + "accept": false, + "reason": "Different model families and different compatibility concerns." + }, + { + "left": "issue:35308", + "right": "issue:42617", + "accept": false, + "reason": "Both involve training/distributed execution, but the concrete bugs are different." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same tokenizer regression: wrong token IDs across models, with 45920 being the broader multi-model report." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes parameter incompatibility is unrelated to distributed build/runtime support." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "Unrelated NVFP4 model runtime issue vs documentation cleanup." + }, + { + "left": "issue:46901", + "right": "issue:47394", + "accept": false, + "reason": "Both are loss-shift bugs, but they are model-specific and describe different failure details; not clearly the same underlying bug." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Different Gemma/Qwen model issues with different symptoms and code paths." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures with no clear shared root cause." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both mention index errors, but they appear to be separate shape/ZeRO-3 problems rather than the same bug." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Different tasks and different runtime failures." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch-version incompatibility is not the same as the PretrainedConfig pydantic regression." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Unable-to-load test failure vs checkpoint corruption; not the same underlying issue." + }, + { + "left": "issue:43541", + "right": "issue:47228", + "accept": false, + "reason": "Both concern model compilation, but they are different models and different regressions." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Unrelated model loading issue vs pydantic config regression." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Different bugs affecting different functionality; no shared code-path issue is apparent." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "ZeRO-3 zero-sized tensor indexing is unrelated to torch/transformers version incompatibility." + }, + { + "left": "issue:43541", + "right": "issue:46869", + "accept": false, + "reason": "Different model/runtime regressions; no concrete shared bug." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading failures and different affected artifacts." + }, + { + "left": "issue:43866", + "right": "issue:46772", + "accept": false, + "reason": "Different compatibility/load issues with no common code-path evidence." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same tokenizer-ID regression reported on different model sets; 45920 is the broader manifestation." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: there are no PRs, and the two accepted soft pairs are the only ones that look plausibly duplicative from the issue titles. The canonical issue choice is conservative and supported by the broader multi-model scope of issue:45920. No clear overstatement of evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 3, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:36010", + "issue:39290", + "issue:39921", + "issue:40990", + "issue:41093", + "issue:41762", + "issue:43054", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43866", + "issue:43872", + "issue:43994", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44479", + "issue:44805", + "issue:44841", + "issue:44918", + "issue:44960", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45362", + "issue:45399", + "issue:45464", + "issue:45507", + "issue:45685", + "issue:45759", + "issue:45800", + "issue:45823", + "issue:45901", + "issue:45902", + "issue:46165", + "issue:46294", + "issue:46640", + "issue:46661", + "issue:46869", + "issue:47228" + ], + "soft_pairs": [ + "issue:45362|issue:45464", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:36010|issue:39290", + "issue:45464|issue:45685", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:44841|issue:45084", + "issue:45507|issue:46869", + "issue:43404|issue:43866", + "issue:43828|issue:45685", + "issue:45399|issue:46661", + "issue:43572|issue:45070", + "issue:41093|issue:41762", + "issue:45800|issue:45823", + "issue:43872|issue:47228", + "issue:43716|issue:47228", + "issue:44805|issue:44918", + "issue:43828|issue:45464", + "issue:43296|issue:43366", + "issue:43366|issue:45464", + "issue:44291|issue:45070", + "issue:43531|issue:45070", + "issue:46869|issue:47228", + "issue:44387|issue:45005", + "issue:45507|issue:47228", + "issue:45020|issue:45800", + "issue:43716|issue:46294", + "issue:44479|issue:46294", + "issue:41093|issue:43596", + "issue:39921|issue:46869", + "issue:43054|issue:43994", + "issue:45759|issue:45902", + "issue:46640|issue:46661", + "issue:43596|issue:46165", + "issue:43716|issue:44918", + "issue:43645|issue:45399", + "issue:39921|issue:45507", + "issue:40990|issue:44960" + ], + "prepared_review_unit_hash": "c3a3b8572b71370165ddafba102cf890cc7476a4b7ee0e62bc59ae00bf3b9062", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 39, + "serialized_chars": 39282, + "estimated_input_tokens": 9821, + "estimated_eval_tokens": 19898 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated Transformers bug reports across model loading, multimodal processors, quantization, vLLM/DeepSpeed, and backend-specific failures. I reject all soft edges; none look like the same concrete bug/change. If forced to name a representative issue, issue:45020 is the broadest regression-style report, but it is only a weak umbrella fit, not a true duplicate hub.", + "confidence": 0.88, + "canonical_issue_reason": "Issue:45020 is the most general regression-style report in the set (recent Transformers versions breaking remote_code models), so it is the best weak umbrella candidate among otherwise unrelated bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue:45020 is the broadest and most reusable representative issue if a single anchor must be chosen, though the cluster is too mixed for a strong canonical duplicate.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Same Qwen3.5 family, but different failures: chat/streaming crash vs chat template API failure. Not the same bug." + }, + { + "left": "issue:44292", + "right": "issue:45901", + "accept": false, + "reason": "Unexpected _is_hf_initialized in init_empty_weights is unrelated to a table-question-answering crash." + }, + { + "left": "issue:39290", + "right": "issue:45823", + "accept": false, + "reason": "Gemma3TextConfig sliding_window_pattern error is a different code path from Gemma4 PLE device mismatch." + }, + { + "left": "issue:36010", + "right": "issue:39290", + "accept": false, + "reason": "ImportError for GenerationMixin and Gemma3TextConfig attr missing are unrelated issues." + }, + { + "left": "issue:45464", + "right": "issue:45685", + "accept": false, + "reason": "Streaming chat failure on Qwen3.5 and an MPS histogram kernel error are unrelated." + }, + { + "left": "issue:43541", + "right": "issue:45685", + "accept": false, + "reason": "Torch Dynamo grouped_mm tracing failure is not the same as the MPS Int histogram error." + }, + { + "left": "issue:43541", + "right": "issue:43828", + "accept": false, + "reason": "Both involve runtime math issues, but grouped_mm tracing and autocast dtype mismatch are different bugs." + }, + { + "left": "issue:44841", + "right": "issue:45084", + "accept": false, + "reason": "Voxtral processor failure and non-template-node compile error are different subsystems and symptoms." + }, + { + "left": "issue:45507", + "right": "issue:46869", + "accept": false, + "reason": "GraniteMoEHybrid invalid method and YOLO multi-GPU device_map failure are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:43866", + "accept": false, + "reason": "lm_head tying bug in Mistral3 and a corrupted Ovis2 checkpoint are not the same issue." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Autocast dtype mismatch and MPS histogram_mps not implemented are unrelated backend bugs." + }, + { + "left": "issue:45399", + "right": "issue:46661", + "accept": false, + "reason": "flash-attn2 fallback logic and remote-code model_class/config_class handling are different code paths." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Both are 5.0 config regressions, but missing pad_token_idx and broken PretrainedConfig pydantic field are distinct bugs." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are IndexErrors, but the mask/tensor shape mismatch is not the same as the zero-size index error under ZeRO-3." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch/Transformers incompatibility and Gemma4 device placement mismatch are separate issues." + }, + { + "left": "issue:43872", + "right": "issue:47228", + "accept": false, + "reason": "bitsandbytes _is_hf_initialized incompatibility and Sam3 bf16 compilation regression are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:47228", + "accept": false, + "reason": "Mistral-3 image preprocessor dtype mismatch and Sam3 bf16 compilation regression do not share the same failure path." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "Similar index/shape wording, but the generic mask mismatch and Qwen3.5 embedding unpacking under TRL are different bugs." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Autocast dtype mismatch is not the same as a Qwen3.5 streaming inference/API failure." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL loading in vLLM and gpt-oss GGUF architecture support are unrelated." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GGUF gpt-oss support and Qwen3.5 streaming chat failure are not the same bug." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "Unexpected _is_hf_initialized in init_empty_weights and a pydantic PretrainedConfig field break are different regressions." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Qwen3-MoE sliding_window handling and PretrainedConfig pydantic issues are unrelated." + }, + { + "left": "issue:46869", + "right": "issue:47228", + "accept": false, + "reason": "YOLO multi-GPU device_map failure and Sam3 bf16 compilation regression are different problems." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "int4 quantization OOM from reserved memory and tied-weights translation-model issues are not the same bug." + }, + { + "left": "issue:45507", + "right": "issue:47228", + "accept": false, + "reason": "GraniteMoEHybrid invalid method and Sam3 bf16 compilation regression do not match." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Broad remote_code breakage and a specific torch 2.4.1/Transformers 5.8.0 incompatibility are related only at a very high level, not the same concrete issue." + }, + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Image-preprocessor dtype mismatch and Qwen3.5 video token/feature mismatch are different multimodal failures." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Both are Qwen video-input regressions, but the affected models and concrete failure modes differ." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Mask-shape mismatch and ZeRO-3 zero-size index error are distinct IndexError classes." + }, + { + "left": "issue:39921", + "right": "issue:46869", + "accept": false, + "reason": "Gemma3N special-token projection error and YOLO multi-GPU device mapping are unrelated." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Siglip2 quality regression and Siglip2 nonsensical AutoModel/pipeline output may be related in theme, but the concrete bug is not clearly the same." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Composite Qwen config unwrapping and uninitialized rotary embedding memory are different code paths." + }, + { + "left": "issue:46640", + "right": "issue:46661", + "accept": false, + "reason": "Flash Attention 2 fallback issue and remote-code config_class AttributeError are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3 BertModel index error and resume-training with local remote-code model are different failures." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Mistral-3 image dtype mismatch and Qwen3.5 embedding unpacking in TRL are separate bugs." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Jupyter custom-model initialization and flash-attn2 fallback gating are unrelated." + }, + { + "left": "issue:39921", + "right": "issue:45507", + "accept": false, + "reason": "Gemma3N special-token projection and GraniteMoEHybrid invalid method are not the same change or bug." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "gpt-oss perplexity anomaly and GLM5 are unrelated issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 4, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:35308", + "issue:36331", + "issue:38175", + "issue:39186", + "issue:39692", + "issue:39921", + "issue:41762", + "issue:42915", + "issue:43054", + "issue:43404", + "issue:43493", + "issue:43644", + "issue:43828", + "issue:43872", + "issue:43873", + "issue:43950", + "issue:43975", + "issue:44292", + "issue:44360", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:45005", + "issue:45183", + "issue:45399", + "issue:45632", + "issue:45698", + "issue:45701", + "issue:45859", + "issue:45902", + "issue:45920", + "issue:45987", + "issue:46032", + "issue:46463", + "issue:46489", + "issue:46491", + "issue:46858", + "issue:46899", + "issue:46932", + "issue:47246", + "issue:47332", + "issue:47333" + ], + "soft_pairs": [ + "issue:42915|issue:45399", + "issue:43404|issue:45005", + "issue:45632|issue:45698", + "issue:39186|issue:44918", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:35308|issue:36331", + "issue:35308|issue:39921", + "issue:44492|issue:44512", + "issue:44530|issue:45005", + "issue:43975|issue:44292", + "issue:45183|issue:46463", + "issue:35308|issue:42915", + "issue:44863|issue:45005", + "issue:39692|issue:43054", + "issue:43950|issue:44534", + "issue:46489|issue:46491", + "issue:36331|issue:41762", + "issue:43644|issue:43950", + "issue:46032|issue:47246", + "issue:43872|issue:44589", + "issue:45920|issue:46489", + "issue:43054|issue:45701", + "issue:46899|issue:46932", + "issue:47332|issue:47333", + "issue:38175|issue:43054", + "issue:43828|issue:44928", + "issue:45987|issue:46858", + "issue:43950|issue:45902", + "issue:44910|issue:45859", + "issue:43054|issue:43493" + ], + "prepared_review_unit_hash": "2e64fbcd869b23d9278f526ea24b2f23f04e237c1f608af415181dd81be96ea2", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37309, + "estimated_input_tokens": 9328, + "estimated_eval_tokens": 18912 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly heterogeneous set, with a few true duplicate subclusters: non-persistent-buffer corruption in v5, input-validation error messaging, cache-length contract violations, Gemma4 quantized embedder dtype casting, Mamba2-family recurrence bugs, and disk-offloaded Qwen3 save failures. The rest are related only at a broad subsystem level and should stay separate.", + "confidence": 0.78, + "canonical_issue_reason": "43950 is the best anchor: it states the broad v5 non-persistent-buffer corruption regression and best subsumes the closer follow-ups in this batch.", + "canonical_pr_reason": null, + "best_issue_reason": "43950 is the most representative issue overall because it has the clearest root-cause framing and the broadest duplicate coverage among the accepted pairs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "Different bugs: one is a FineGrainedFP8Config failure, the other is a flash-attn2 fallback gating issue." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Related to tied weights, but the model families and concrete failure modes differ too much to treat as the same bug." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both involve custom-code loading, but one is a leaf-directory cache collision and the other is wrong module selection after save_pretrained." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP 2-D weight shape and TRL embedding unpacking are different code paths and symptoms." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is not the same as tied-weight regression." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "A DSA indexer discussion and a removed docs command are unrelated." + }, + { + "left": "issue:35308", + "right": "issue:36331", + "accept": false, + "reason": "IterableDataset multi-GPU crash is unrelated to the custom compute_loss signature break." + }, + { + "left": "issue:35308", + "right": "issue:39921", + "accept": false, + "reason": "No shared underlying failure; training crash vs Gemma3N projection/tokenizer error." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "Just a typo vs a docs/regression issue." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache/linear_attention crash is unrelated to tied weights." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Tokenizer detokenization regression and NVFP4 runtime error are different problems." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": true, + "reason": "46463 looks like the T5-specific manifestation of the broader XOR validation/error-message bug described in 45183." + }, + { + "left": "issue:35308", + "right": "issue:42915", + "accept": false, + "reason": "Unrelated training-distribution crash vs FP8 configuration failure." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "Model loading failure is not the same as tied-weight translation-model regression." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Docs example errors are unrelated to SigLIP2 embedding quality/discrepancy reports." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same v5 non-persistent-buffer corruption regression; 44534 is a near-duplicate symptom report." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but on different models and different failure modes." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Trainer API signature break vs DeepSpeed ZeRO-3 loading failure are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "Same non-persistent-buffer junk/corruption regression in transformers 5.x." + }, + { + "left": "issue:46032", + "right": "issue:47246", + "accept": true, + "reason": "Both describe the same Mamba2-family recurrence/use_cache bug; 47246 is the NemotronH port that missed the earlier fix." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Different quantization backend incompatibilities with different exception types." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Tokenizer-output regressions, but they affect different tokenizers/models and are not the same bug." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "SigLIP2 embedding quality and version-dependent tokenization are unrelated." + }, + { + "left": "issue:46899", + "right": "issue:46932", + "accept": true, + "reason": "46932 explicitly says it is the same issue; both point to the unsafe dtype cast in Gemma4 quantized multimodal embedders." + }, + { + "left": "issue:47332", + "right": "issue:47333", + "accept": true, + "reason": "Same disk-offloaded Qwen3 save failure, just on different Qwen3 model variants." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Different SigLIP2 symptoms: zero probabilities vs worse text embeddings." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Autocast dtype mismatch is not the same as the Qwen3.5 SDPA/NaN RLHF issue." + }, + { + "left": "issue:45987", + "right": "issue:46858", + "accept": true, + "reason": "Same get_seq_length contract bug; 46858 is the backend-specific torch.compile manifestation." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": true, + "reason": "Same v5 non-persistent-buffer/materialization regression; 45902 is a specific buffer-level manifestation." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Flash-attention packed-sequence crash and CPU-offload rotary-embedding incompatibility are different defects." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Both concern SigLIP2, but one is an embedding-quality complaint and the other is an HF-vs-JAX implementation discrepancy." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays mostly conservative. The accepted pairs are plausible single-bug subclusters, especially the non-persistent-buffer regression around #43950, the cache-length contract issues, the Gemma4 dtype-cast pair, the Mamba2/NemotronH recurrence bug, and the Qwen3 disk-offload save failures. The rejected pairs are generally framed narrowly enough to avoid overclaiming. No soft-edge verdict appears to overstate the evidence materially." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 5, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:24643", + "issue:28530", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43643", + "issue:43782", + "issue:43792", + "issue:43824", + "issue:43827", + "issue:43828", + "issue:43856", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44220", + "issue:44368", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44509", + "issue:44661", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45237", + "issue:45305", + "issue:45356", + "issue:45399", + "issue:45488", + "issue:45663", + "issue:45685", + "issue:45715", + "issue:45759", + "issue:45823", + "issue:45902", + "issue:46315", + "issue:46489", + "issue:46661", + "issue:46869" + ], + "soft_pairs": [ + "issue:45663|issue:45902", + "issue:42947|issue:44387", + "issue:44403|issue:44451", + "issue:24643|issue:28530", + "issue:43975|issue:45237", + "issue:41762|issue:42915", + "issue:44403|issue:44488", + "issue:43643|issue:45020", + "issue:44991|issue:45399", + "issue:42947|issue:43856", + "issue:45685|issue:45823", + "issue:43824|issue:43872", + "issue:45081|issue:45399", + "issue:45020|issue:45823", + "issue:43782|issue:44960", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:41553|issue:43883", + "issue:45020|issue:45356", + "issue:43872|issue:46869", + "issue:42491|issue:44368", + "issue:45759|issue:46661", + "issue:41553|issue:43881", + "issue:45020|issue:46661", + "issue:45488|issue:46489", + "issue:43404|issue:43950", + "issue:45715|issue:46315", + "issue:43827|issue:44509", + "issue:43881|issue:43883", + "issue:43257|issue:45305", + "issue:43792|issue:44220" + ], + "prepared_review_unit_hash": "926cef98dec0e51e799188fd6ee0df424eeda13aad5032f9df0346af730661ca", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37412, + "estimated_input_tokens": 9353, + "estimated_eval_tokens": 18962 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose collection of Transformers v5-era regressions around model/tokenizer loading, with only a couple of clear duplicate-like pairs. The strongest canonical issue is the broad remote-code loading regression; the only soft edges I\u2019d accept are the docs pipeline-removal pair and the Whisper/fbank feature-extraction pair.", + "confidence": 0.52, + "canonical_issue_reason": "issue:45020 is the broadest user-facing regression here, centered on remote-code model loading and general enough to sit above the more specific loading/tokenizer failures.", + "canonical_pr_reason": null, + "best_issue_reason": "It has the widest scope and best fits the cluster\u2019s dominant theme of v5 loading regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Different bugs: Gemma4 FSDP/shared_kv_states error vs Qwen3_5 MoE inv_freq materialization bug." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Both are memory/efficiency complaints, but they hit different mechanisms: LoRA checkpointing vs int4 reserved memory/OOM." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic loading noise vs a specific model load failure; no shared concrete code path." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "Unrelated training/deepspeed runtime error vs early-stopping metric-selection issue." + }, + { + "left": "issue:43975", + "right": "issue:45237", + "accept": false, + "reason": "Tokenizer detokenization bug vs AMD GPU model-support failure; different subsystems." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Different DeepSpeed/Qwen problems: ZeRO-3 load crash vs FineGrainedFP8Config failure." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Both are generic loading failures, but they are model-specific and not the same bug." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Remote-code is the common theme, but one is missing fields from AutoConfig and the other is a broad loading regression umbrella." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer-loading failure for a specific model vs flash-attn fallback gating; unrelated." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "Different symptoms and code paths: gradient checkpointing vs MoE training memory usage." + }, + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "MPS int histogram limitation is unrelated to Gemma4 device-mapping mismatch." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Import error for a model class vs bitsandbytes Int8Params incompatibility; not the same bug." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer regex crash vs flash-attn fallback blocking; different failure points." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Broad remote-code loading regressions vs Gemma4 PLE device mismatch; different concrete problems." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight-only loading error vs a generic GLM5 issue title; insufficient evidence of the same bug." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "Tokenizer mapping-name edge case vs broad remote-code regressions; not the same concrete fix." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Autocast dtype mismatch vs Gemma4 PLE device mismatch; different code paths." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Bad error message for AutoTokenizer loading Voxtral vs missing all_tied_weights_keys attribute; unrelated." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Remote-code loading regression umbrella vs Kimi tokenizer codec-handling regression; different tokenizer bug." + }, + { + "left": "issue:43872", + "right": "issue:46869", + "accept": false, + "reason": "bitsandbytes constructor incompatibility vs YOLO multi-GPU device-map failure; unrelated." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "LoRA compatibility break across versions vs a warning about tie_word_embeddings; not the same underlying bug." + }, + { + "left": "issue:45759", + "right": "issue:46661", + "accept": false, + "reason": "Composite multimodal config unwrapping bug vs remote-code model_class config_class AttributeError; different failure modes." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Different loading issues for different models; only superficial similarity in being model-load failures." + }, + { + "left": "issue:45020", + "right": "issue:46661", + "accept": false, + "reason": "The latter is a specific remote-code AttributeError, but not clearly the same concrete bug as the broad regression umbrella." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Both involve tokenizer regressions, but they affect different DeepSeek families and appear to need separate fixes." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "lm_head tying issue vs non-persistent buffer corruption; distinct model-loading bugs." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "One is skip_special_tokens per-id loop inefficiency; the other is convert_tokens_to_ids slowdown with many added tokens." + }, + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Both are the same docs cleanup for removed v5 pipeline tasks; title wording differs but the underlying change is identical." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "Different model-specific loading failures with no shared code-path evidence." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Qwen3 MoE weight conversion on accelerate+deepspeed vs gradient averaging with GAS; not the same bug." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": true, + "reason": "Both point to the same audio feature-extraction path; the Whisper failure and the _torch_extract_fbank_features() issue look like the same underlying bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative: it identifies the cluster as a loose mix of v5-era loading/tokenizer regressions, picks a plausible broad canonical issue (issue:45020), and only accepts soft pairs where the titles strongly support a shared fix (docs pipeline removals and the Whisper/fbank audio path). I don\u2019t see an overstatement or an unsafe duplicate claim here." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 6, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36010", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42617", + "issue:42915", + "issue:42947", + "issue:43366", + "issue:43531", + "issue:43596", + "issue:43638", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43824", + "issue:43828", + "issue:43854", + "issue:43940", + "issue:43950", + "issue:43957", + "issue:44155", + "issue:44479", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44683", + "issue:44805", + "issue:44863", + "issue:44918", + "issue:44928", + "issue:44991", + "issue:45313", + "issue:45325", + "issue:45584", + "issue:45588", + "issue:45663", + "issue:45685", + "issue:45823", + "issue:45834", + "issue:45902", + "issue:45923", + "issue:46402", + "issue:46468", + "issue:46515", + "issue:46894", + "issue:46897" + ], + "soft_pairs": [ + "issue:46894|issue:46897", + "issue:36331|issue:41093", + "issue:42947|issue:45923", + "issue:43940|issue:45313", + "issue:24643|issue:30064", + "issue:44479|issue:44560", + "issue:43854|issue:44863", + "issue:41093|issue:44805", + "issue:43366|issue:43828", + "issue:43716|issue:46468", + "issue:44928|issue:45588", + "issue:43366|issue:43531", + "issue:44530|issue:44863", + "issue:43596|issue:46515", + "issue:44863|issue:46402", + "issue:45325|issue:45902", + "issue:43366|issue:44589", + "issue:43638|issue:46515", + "issue:43716|issue:44560", + "issue:45663|issue:45823", + "issue:41553|issue:44991", + "issue:42915|issue:43957", + "issue:44155|issue:45584", + "issue:43824|issue:43957", + "issue:42617|issue:43366", + "issue:43950|issue:45834", + "issue:44560|issue:45685", + "issue:44560|issue:44918", + "issue:36010|issue:42915", + "issue:44683|issue:46515", + "issue:43645|issue:43646" + ], + "prepared_review_unit_hash": "9b6f4d1cad3e6263af6e6530352fba0b95956b9a7e46efb30934ed6282e2af0f", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37079, + "estimated_input_tokens": 9270, + "estimated_eval_tokens": 18796 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The supported duplicates are the two training-loss double-shift reports, the exact mask/tensor shape mismatch reports, and the custom-model-init regression pair. The DeepSpeed/Bert ZeRO-3 reports are not accepted here; the packet does not show a sufficiently shared underlying bug to merge them. Most other soft edges look like title-similarity false positives.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43646 is the cleanest canonical representative for the accepted custom-model-initialization regression: it states the broad Transformers 5.0.0 breakage without the notebook-specific framing in 43645.", + "canonical_pr_reason": null, + "best_issue_reason": "Among the accepted issue duplicates, 43646 is the most general and concise issue to anchor the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46894", + "right": "issue:46897", + "accept": true, + "reason": "Same training-loss double-shift bug pattern across two models; strong duplicate match." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Different failures: trainer compute_loss signature regression vs mask/tensor shape mismatch." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "Related training features, but different bugs: LoRA checkpointing behavior vs missing model flag." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both involve ZeRO-3 weight loading, but the model families and failure modes differ too much to merge." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Unrelated: DeepSpeed training error vs image processor void-segmentation handling." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are Qwen video regressions, but the concrete symptoms and code paths differ." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different model-loading problems for different models; not the same defect." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": true, + "reason": "Near-identical mask-shape mismatch IndexError; same underlying bug and code path." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF architecture support request vs autocast dtype mismatch are unrelated." + }, + { + "left": "issue:43716", + "right": "issue:46468", + "accept": false, + "reason": "Different model families and different loading/forward symptoms." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "RLHF NaNs from 3D position ids are unrelated to the flash-attention s_aux=None crash." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "Different Qwen/GGUF feature and sliding-window issues; not the same bug." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "PagedAttentionCache group-type crash and NemotronH checkpoint loading are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:46515", + "accept": false, + "reason": "Specific DeepSpeed/Bert Zero-3 index error vs a broad torch incompatibility report." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Different checkpoint-loading failures for different model families." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Different Qwen rope/position-id regressions affecting different paths." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "GGUF support request and Float8 storage lookup failure are unrelated." + }, + { + "left": "issue:43638", + "right": "issue:46515", + "accept": false, + "reason": "A specific Bert/ZeRO-3 index error is not the same as a generic torch-version incompatibility." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different model families and different video/image processing failures." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states KeyError and device-map mismatch are different bugs." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Different tokenizer-loading failures for different models/repos." + }, + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Qwen3Moe FP8 config failure is unrelated to meta-device loading regressions." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "AudioFlamingo3 batch leakage and Whisper empty-transcription handling are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Specific import error for one class vs a meta-device loading problem." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py runtime failure and GGUF gpt-oss support are unrelated." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer corruption is a different regression from long-OCR index errors." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Qwen video StopIteration and MPS histogram Int support are unrelated." + }, + { + "left": "issue:44560", + "right": "issue:44918", + "accept": false, + "reason": "Qwen video input failure and TRL SFT unpacking failure affect different code paths." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "GenerationMixin import failure and Qwen3Moe FP8 config failure are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:46515", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 and a generic torch 2.4.1 incompatibility report are not the same bug." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same custom-model-initialization regression; 43645 is the notebook-specific manifestation of 43646." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The output is mostly conservative, but the summary overstates the evidence: it says the deepspeed/Bert zero-3 reports are among the 'clearest duplicates,' yet both relevant verdicts in the list are marked false. That inconsistency makes the summary insufficiently grounded in the packet. The accepted pairs that are actually supported are the two training-loss double-shift reports, the exact mask-shape mismatch reports, and the custom-model-init reports." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 7, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41553", + "issue:41720", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43317", + "issue:43550", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43644", + "issue:43720", + "issue:43994", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44560", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44960", + "issue:44977", + "issue:44991", + "issue:45030", + "issue:45276", + "issue:45313", + "issue:45357", + "issue:45507", + "issue:45672", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:46459", + "issue:46468", + "issue:46489", + "issue:46512", + "issue:46515", + "issue:46846", + "issue:46858", + "issue:46869", + "issue:46899", + "issue:47254" + ], + "soft_pairs": [ + "issue:44910|issue:46858", + "issue:43550|issue:44910", + "issue:41553|issue:44488", + "issue:45823|issue:46515", + "issue:42947|issue:47254", + "issue:46468|issue:46869", + "issue:45812|issue:46489", + "issue:44560|issue:44805", + "issue:44661|issue:45823", + "issue:44877|issue:45030", + "issue:38175|issue:43994", + "issue:43638|issue:44661", + "issue:44661|issue:44805", + "issue:43596|issue:44661", + "issue:42491|issue:43720", + "issue:45672|issue:46899", + "issue:43632|issue:45834", + "issue:41720|issue:42915", + "issue:43317|issue:44756", + "issue:42915|issue:44756", + "issue:38175|issue:39692", + "issue:43644|issue:44534", + "issue:42491|issue:45276", + "issue:44960|issue:44977", + "issue:44451|issue:44488", + "issue:45507|issue:46468", + "issue:45313|issue:45357", + "issue:44863|issue:46846", + "issue:46459|issue:46512", + "issue:39692|issue:43994", + "issue:44488|issue:44991" + ], + "prepared_review_unit_hash": "a7acb913d4bb5d06e282bf2c7a707234764bf44e4f44e3f2aa4a53ebb361e71c", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36340, + "estimated_input_tokens": 9085, + "estimated_eval_tokens": 18426 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous with a few clear duplicate pairs. I moved the canonical issue anchor away from 44488 to the higher-discussion loading report 41553, and I downgraded the SigLIP2 pair to a non-duplicate because the evidence shows similarity but not the same root cause.", + "confidence": 0.74, + "canonical_issue_reason": "Issue 41553 has the strongest discussion signal in the packet and is a cleaner representative of the loading/error-report subset than the more ambiguous model-specific cases.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 41553 is the best global issue anchor here because it is the most discussion-heavy and broadly framed, even though the overall cluster still contains several unrelated model-specific bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44910", + "right": "issue:46858", + "accept": false, + "reason": "Different bugs: Qwen3.5 flash-attn/3D position_ids vs static-cache slice arithmetic on non-CUDA backends." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Same broad attention/compile area, but different models, triggers, and failure paths." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but they involve different models and no evidence of the same root cause." + }, + { + "left": "issue:45823", + "right": "issue:46515", + "accept": false, + "reason": "Device-mapping forward mismatch vs a torch-version incompatibility; unrelated symptoms." + }, + { + "left": "issue:42947", + "right": "issue:47254", + "accept": false, + "reason": "Both mention PEFT/checkpointing, but one is ineffective checkpointing and the other is a ZeRO-3 checkpoint error." + }, + { + "left": "issue:46468", + "right": "issue:46869", + "accept": false, + "reason": "Different models and contexts: Nemotron example vs YOLO multi-GPU device_map loading." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but for different model families and likely separate code paths." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different failures on different inputs/models; only a vague runtime-error similarity." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Tokenizer-mapping helper failure vs Gemma4 device mismatch during forward." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Both are config/load issues, but they affect different model types and are not the same bug." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": false, + "reason": "Related SigLIP2 symptoms, but the packet does not establish the same underlying bug; similarity is only at the model/symptom level." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Different code paths: ZeRO-3 Bert initialization vs tokenizer-mapping behavior." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer-mapping helper bug vs a mask/tensor shape IndexError; not the same issue." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Different failures: Deepspeed Zero-3 Bert init vs tokenizer-mapping support." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Both loading-related, but one is a Qwen3 MoE LoRA incompatibility and the other is BitNet packed-weight handling." + }, + { + "left": "issue:45672", + "right": "issue:46899", + "accept": true, + "reason": "Same Gemma4 multimodal 4-bit quantization bug: unsafe dtype handling in embedders/audio-vision paths." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "Transformers v5 flag regression vs a Kosmos2.5 OCR index error; unrelated." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Different Qwen-family failures with different mechanisms: cuda assert on device mapping vs FineGrainedFP8Config." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Both are memory/loading problems, but device_map offload failure and Strix Halo mmap OOM are distinct bugs." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "No meaningful overlap beyond general memory/perf symptoms." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Same model family, but a docs/example mismatch with quantization failure is not the same as the zero-probabilities inference bug." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same exact v5 regression about non-persistent buffers being filled with junk; phrasing differs, bug is the same." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "Different topics: Qwen3 MoE LoRA portability vs Gemma4 embedding resize behavior." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different models and failure modes; both are generation issues but not the same code-path problem." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Both are tokenizer-loading complaints, but for different models with no evidence of the same root cause." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "Different models and unrelated failures." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "Weight-loading ZeRO-3 issue vs save_pretrained visual-encoder key regression; not the same bug." + }, + { + "left": "issue:44863", + "right": "issue:46846", + "accept": false, + "reason": "NemotronH checkpoint loading vs Qwen3.5 tensor-parallel plan errors; unrelated." + }, + { + "left": "issue:46459", + "right": "issue:46512", + "accept": false, + "reason": "Both involve dtype, but one is ignored dtype during checkpoint loading and the other is stale config.dtype after from_config." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but docs-example mistakes and actual inference nonsense are not clearly the same underlying bug." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-load regressions on v5, but the affected models and failure details differ too much to merge confidently." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The cluster summary is mostly grounded, but two parts overstate evidence. First, the canonical-issue rationale for issue:44488 is not well supported: it does not have the strongest discussion signal in the packet, and its inbound references are not uniquely high compared with other candidates. Second, the accepted soft edge issue:38175 \u2194 issue:43994 is too strong for the evidence shown: both are SigLIP2-related and symptomatically similar, but the packet does not clearly establish the same underlying bug, so that should be treated more conservatively. The other accepted duplicates look more defensible." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 8, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42947", + "issue:43257", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:44075", + "issue:44079", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44387", + "issue:44402", + "issue:44479", + "issue:44488", + "issue:44756", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44960", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45362", + "issue:45406", + "issue:45488", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:45859", + "issue:46050", + "issue:46468", + "issue:46829", + "issue:46897", + "issue:46901" + ], + "soft_pairs": [ + "issue:43257|issue:43854", + "issue:43644|issue:43645", + "issue:43856|issue:43873", + "issue:42491|issue:45335", + "issue:42491|issue:45292", + "issue:44164|issue:44291", + "issue:43856|issue:44387", + "issue:43854|issue:43931", + "issue:45488|issue:45812", + "issue:44960|issue:46050", + "issue:43708|issue:44075", + "issue:45276|issue:45335", + "issue:43425|issue:43957", + "issue:44291|issue:44756", + "issue:45362|issue:45406", + "issue:41720|issue:44155", + "issue:45663|issue:45759", + "issue:39692|issue:44960", + "issue:44910|issue:45588", + "issue:44402|issue:44488", + "issue:46897|issue:46901", + "issue:43873|issue:44387", + "issue:43404|issue:43454", + "issue:42947|issue:45704", + "issue:45663|issue:45859", + "issue:44155|issue:45823", + "issue:43957|issue:44756", + "issue:44079|issue:46829", + "issue:44863|issue:46468", + "issue:44479|issue:44877" + ], + "prepared_review_unit_hash": "1829bac9ac2c6ebfae6a801f90bd751af13d9f650bd545555bf3016f6829ad46", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36843, + "estimated_input_tokens": 9211, + "estimated_eval_tokens": 18678 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: most items are distinct model-loading, quantization, or training bugs. The only clear duplicate-style match is the Gemma4/t5gemma `resize_token_embeddings` propagation issue; everything else is too model- or code-path-specific to merge.", + "confidence": 0.73, + "canonical_issue_reason": "issue:45276 is the strongest canonical issue among the true duplicates: it is the more detailed report of the `resize_token_embeddings` propagation bug, and issue:45335 reads like a narrower variant of the same underlying problem.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best overall issue candidate for the set: it has the most discussion/inbound references and is a broad, actively relevant quantization/offloading report, while most other items are isolated model-specific bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Both are Qwen3/GLM loading failures, but one is a conversion/offload problem and the other is a specific unit-test load mismatch; different failure modes and fixes." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Both mention Transformers 5 regressions, but one is about junk-filled non-persistent buffers and the other about Jupyter/custom-model initialization; not the same bug." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both involve Qwen3 MoE and memory, but one is general training-memory inefficiency and the other is offloading behavior with quantization; too different to dedupe." + }, + { + "left": "issue:42491", + "right": "issue:45335", + "accept": false, + "reason": "LoRA hf4.x/hf5.x incompatibility for Qwen3 MoE is unrelated to a Gemma/T5Gemma embedding-resize bug." + }, + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Version incompatibility for a Qwen3 MoE LoRA model is not the same underlying issue as `resize_token_embeddings` failing to update embeddings." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "One is broken `extra_state` save/load handling, the other is `init_empty_weights` rejecting an unexpected `_is_hf_initialized` argument; different code paths." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both are memory-related, but one is training inefficiency in Qwen3 MoE and the other is int4 quantization increasing reserved CUDA memory/OOM; not the same fix." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Both are Qwen3-VL loading problems, but one is a unit-test/model-loading failure and the other is a weight-shape mismatch for a specific checkpoint." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Both affect Granite-related tokenization, but one is a pre-tokenizer override regression and the other is incorrect token IDs; related subsystem, different bug." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "GLM5 is too vague to match the specific sinq save/load breakage; no evidence they are the same underlying issue." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint-resume step calculation and optimizer SGD args being ignored are unrelated." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": true, + "reason": "Both report `resize_token_embeddings` failing to propagate to model-specific embeddings; the second is a narrower manifestation of the same bug pattern." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 compatibility is not the same as meta-device loading failures on Transformers 5." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "`init_empty_weights` argument handling and disabling mmap on Strix Halo are unrelated issues." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Qwen3.5 chat crashes and Gemma4Processor missing `_tokenizer` are different model/processor bugs." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "A800 `cudaErrorAssert` with Qwen3 auto device mapping is unrelated to AudioFlamingo3 batched inference token/embedding leakage." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "FSDP2 `shared_kv_states` rebuilding and `AutoModelForCausalLM.from_config` not unwrapping `text_config` are different composite-model issues." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "SigLIP2 documentation/example errors are unrelated to GLM5 loading behavior." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both crash in flash-attention contexts, but one is a packed-sequence misinterpretation from 3D `position_ids` and the other is an `s_aux=None` sink-less-model error." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "A tokenizer vocab-size mismatch for one model is not the same as a broader `cjvt/sleng-bert` loading failure." + }, + { + "left": "issue:46897", + "right": "issue:46901", + "accept": false, + "reason": "Both describe a double-shift training-loss pattern, but they are different model implementations and there is no evidence of one shared code path or one mergeable fix." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both involve quantization and memory, but one is offloading behavior and the other is increased reserved memory under int4 quantization; not the same bug." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both mention untied `lm_head` weights, but they are separate model families and likely separate tie-weight implementations." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing ineffectiveness with PEFT LoRA is unrelated to an apex RMSNorm memory leak." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Gemma-4 FSDP2 key errors and Qwen3.5-MoE CPU-offload incompatibility are distinct issues." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo3 batch leakage and Gemma4 PLE device mismatch are different bugs in different model paths." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading regressions and Strix Halo mmap OOM are unrelated." + }, + { + "left": "issue:44079", + "right": "issue:46829", + "accept": false, + "reason": "`ModelOutput` key assignment and VideoPrism hidden-state field wiring are different output-structure bugs." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint loading failure and the official Nemotron example being broken are related only by model family, not the same concrete issue." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video-input regression for Qwen VL models and strict config blocking `granite_speech` loading are separate bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet and stays conservative. The best-issue choice is supported by discussion/inbound counts, and the only accepted soft-edge pair (issue:45276 vs issue:45335) is a plausible same-bug duplicate within the resize_token_embeddings propagation family. The remaining pairs are reasonably rejected as model- or code-path-specific rather than clearly mergeable duplicates." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 9, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39921", + "issue:42796", + "issue:43335", + "issue:43381", + "issue:43425", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43828", + "issue:43854", + "issue:43957", + "issue:44164", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44530", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44912", + "issue:44928", + "issue:45084", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45698", + "issue:45701", + "issue:45812", + "issue:45823", + "issue:46315", + "issue:46326", + "issue:46491", + "issue:46869", + "issue:46894", + "issue:46901", + "issue:47228", + "issue:47332", + "issue:47448" + ], + "soft_pairs": [ + "issue:43335|issue:45663", + "issue:39921|issue:43425", + "issue:44387|issue:44756", + "issue:44292|issue:44912", + "issue:36010|issue:39404", + "issue:45276|issue:45292", + "issue:44164|issue:44756", + "issue:43957|issue:47332", + "issue:43381|issue:44928", + "issue:45588|issue:45663", + "issue:43645|issue:44164", + "issue:43425|issue:47228", + "issue:42796|issue:46326", + "issue:44589|issue:44756", + "issue:43425|issue:43828", + "issue:45292|issue:45335", + "issue:43854|issue:45216", + "issue:43646|issue:43708", + "issue:44402|issue:44451", + "issue:45684|issue:45698", + "issue:43425|issue:46869", + "issue:46894|issue:46901", + "issue:44530|issue:45084", + "issue:45812|issue:46315", + "issue:46315|issue:46491", + "issue:44683|issue:45823", + "issue:39404|issue:45701", + "issue:43611|issue:43645", + "issue:44479|issue:44805", + "issue:44589|issue:45084", + "issue:45823|issue:47448" + ], + "prepared_review_unit_hash": "429467ae9ec251cbd2abb8c2471f6468434ae86de09d43cec1be92e655b70753", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 36976, + "estimated_input_tokens": 9244, + "estimated_eval_tokens": 18744 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most candidate pairs are false positives across unrelated model-, tokenizer-, or runtime-specific regressions. The only clear duplicate cluster is the resize_token_embeddings bug: issue 45292 is the generic report, while 45276 and 45335 are model-specific manifestations.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45292 most cleanly states the shared underlying regression: `resize_token_embeddings` fails to propagate updates to output embeddings. It is the broadest representative of the accepted duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45292 is the best issue candidate because it describes the core bug generically, without narrowing it to one architecture, while the other accepted issues are specific instances of the same failure mode.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "Different models and different failures: sparse-layer construction vs Gemma4 FSDP attention state rebuilding." + }, + { + "left": "issue:39921", + "right": "issue:43425", + "accept": false, + "reason": "Special-token projection error in Gemma3N is unrelated to a Torch 2.10 compatibility issue." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention OOM, but the causes differ: int4 quantization memory growth vs disabling mmap on Strix Halo." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "Different quantization paths and symptoms (`NVFP4` run failure vs `MXFP4` fallback to bf16) on different models." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for `GenerationMixin` is unrelated to Whisper pipeline `return_language` behavior." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": true, + "reason": "Same underlying `resize_token_embeddings` propagation bug; 45276 is a Gemma4-specific manifestation of the generic output-embedding issue in 45292." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "`extra_state` serialization failure and mmap-induced OOM are different save/load problems." + }, + { + "left": "issue:43957", + "right": "issue:47332", + "accept": false, + "reason": "Meta-device loading regressions and disk-offloaded save failures are different code paths." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Gradient checkpointing in eval mode is unrelated to RLHF NaNs from 3D position ids and SDPA fallback." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Flash-attention sink-less `s_aux=None` crash is not the same bug as Gemma4 FSDP2 shared KV state rebuilding." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Custom-model notebook initialization regression is different from `save/from_pretrained` extra_state handling." + }, + { + "left": "issue:43425", + "right": "issue:47228", + "accept": false, + "reason": "Both mention torch/version sensitivity, but one is broad incompatibility and the other is a specific bf16 compilation regression." + }, + { + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer BPE-marker handling and `continue_final_message` behavior are different tokenizer bugs." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage deserialization error is unrelated to the Strix Halo mmap OOM issue." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch 2.10 incompatibility is not the same as Phi-tiny-MoE autocast dtype mismatch." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same `resize_token_embeddings` regression, with 45335 showing the decoder.embed_tokens variant of the output-embedding propagation failure." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "Different model-loading/save regressions on different models and APIs." + }, + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Custom model initialization regression is unrelated to Trainer `resume_from_checkpoint` step \u0570\u0561\u0577\u057eculation." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "The ScandiBERT load failure may be related in theme, but the titles do not establish the same special-token/vocab bug." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Read-only permission propagation from `save_pretrained` is different from loading the wrong custom module after save." + }, + { + "left": "issue:43425", + "right": "issue:46869", + "accept": false, + "reason": "Torch version incompatibility and multi-GPU `device_map=\"auto\"` failure are different issues." + }, + { + "left": "issue:46894", + "right": "issue:46901", + "accept": false, + "reason": "Both are training-loss shift bugs, but they are model-specific reports without enough evidence of the same code path." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache group-type failure is unrelated to the compiler error about non-template nodes." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Wrong token IDs for Granite models and slow `convert_tokens_to_ids` are different tokenizer regressions." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Performance regression in token conversion and CodeLlama spacing round-trip bug are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch >= 2.9 and Gemma4 device-mapping mismatch are different failures." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper pipeline regression is unrelated to a general tokenization change report." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Base-model-prefix loading breakage and Jupyter custom-model initialization breakage are not the same concrete bug." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression for Qwen VL models is unrelated to the mask/tensor shape mismatch." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Float8 storage lookup failure is not the same as the compiler rejecting non-template nodes." + }, + { + "left": "issue:45823", + "right": "issue:47448", + "accept": false, + "reason": "Gemma4 device mismatch during forward is unrelated to the model-architecture recognition failure." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate cluster is the resize_token_embeddings/output_embeddings trio (45292 with 45276 and 45335), and the other pair rejections are appropriately cautious given only the issue titles." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 10, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:39921", + "issue:40990", + "issue:42796", + "issue:42915", + "issue:43404", + "issue:43452", + "issue:43541", + "issue:43596", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43716", + "issue:43746", + "issue:43856", + "issue:43901", + "issue:43906", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "issue:44462", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44671", + "issue:44683", + "issue:44756", + "issue:44898", + "issue:44912", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45479", + "issue:45759", + "issue:45823", + "issue:46050", + "issue:46326", + "issue:46489", + "issue:46665", + "issue:46821", + "issue:46894", + "issue:46899", + "issue:47050", + "issue:47331", + "issue:47333" + ], + "soft_pairs": [ + "issue:44164|issue:45759", + "issue:43994|issue:44079", + "issue:40990|issue:47331", + "issue:44265|issue:44898", + "issue:36032|issue:43452", + "issue:39404|issue:46665", + "issue:43901|issue:44509", + "issue:44991|issue:45081", + "issue:46050|issue:46326", + "issue:44671|issue:44991", + "issue:43906|issue:46326", + "issue:43906|issue:44530", + "issue:39404|issue:44912", + "issue:45823|issue:46899", + "issue:44164|issue:46821", + "issue:44756|issue:47333", + "issue:43596|issue:44683", + "issue:43856|issue:44756", + "issue:43632|issue:47050", + "issue:44488|issue:44987", + "issue:36032|issue:39921", + "issue:36032|issue:42915", + "issue:44488|issue:45081", + "issue:43716|issue:45084", + "issue:42796|issue:43404", + "issue:43746|issue:47331", + "issue:43611|issue:43644", + "issue:43541|issue:45084", + "issue:44462|issue:46489", + "issue:45479|issue:46894" + ], + "prepared_review_unit_hash": "9b9466877ed34d99778f64017fa125a23da06153b4fa67f42c0ce524617622a3", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37001, + "estimated_input_tokens": 9251, + "estimated_eval_tokens": 18758 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is noisy and mostly heterogeneous: many items are Transformers v5 regressions, but they span different code paths (tokenizer loading, model loading, quantization, docs, generation, export, and model outputs). The only clearly close pairings are still not strong enough to treat as duplicates.", + "confidence": 0.33, + "canonical_issue_reason": "issue:43611 is the broadest and most central v5 loading-regression report in the set; it best matches the cluster\u2019s recurring theme of model/tokenizer loading breakage, even though the overall cluster is not a clean duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43611 is the best single representative anchor for triage because it describes a concrete, high-level loading failure in Transformers v5 and is closer to the shared pattern than the more niche reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "Different failures: extra_state save/load handling vs. from_config text_config unwrapping for multimodal Qwen configs." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Different bug classes: wrong SIGLIP2 outputs vs. ModelOutput key assignment when a key was previously None." + }, + { + "left": "issue:40990", + "right": "issue:47331", + "accept": false, + "reason": "Unrelated problem surfaces: perplexity evaluation anomaly vs. Metal quantization batched generate corruption." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "Both are export/vision-related, but the concrete code paths differ: torch.export with torch_compilable_check vs. Perceiver interpolation handling." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "Both involve loading, but one is a T5Tokenizer add_special_tokens conflict and the other is gguf_file resolution for AutoTokenizer/AutoModel." + }, + { + "left": "issue:39404", + "right": "issue:46665", + "accept": false, + "reason": "Whisper pipeline return_language regression is unrelated to inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs issues, but they concern different removed/deprecated pipeline/task references." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-loading bugs, but one is a repo-specific tokenizer load failure and the other is a Mistral regex patch crash in a distinct helper path." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantized save/load breakage is not the same as continue_final_message behavior for mistral_common tokenizers." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT MLM prediction regression and est-roberta tokenizer loading failure are different model behaviors and different code paths." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "A reproduction of an external issue is not the same as a Mistral tokenizer continue_final_message bug." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Different subsystems: reproduction of a prior bug vs. PagedAttentionCache linear_attention crash on Qwen3.5." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline language-return behavior is unrelated to gpt-oss MXFP4 quantization fallback." + }, + { + "left": "issue:45823", + "right": "issue:46899", + "accept": false, + "reason": "Both mention multimodal dtype/device issues, but one is Gemma4 PLE device_map forwarding and the other is unsafe dtype casting in quantized embedders." + }, + { + "left": "issue:44164", + "right": "issue:46821", + "accept": false, + "reason": "save/from_pretrained extra_state handling is unrelated to a TimesFM window_size AttributeError." + }, + { + "left": "issue:44756", + "right": "issue:47333", + "accept": false, + "reason": "Both are resource/offload related at a high level, but one is mmap OOM on Strix Halo and the other is disk-offloaded Qwen save failure." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Zero3/BertModel indexing failure is a different issue than compiled flex_attention failing on torch>=2.9." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory inefficiency is not the same as mmap-induced OOM on Strix Halo." + }, + { + "left": "issue:43632", + "right": "issue:47050", + "accept": false, + "reason": "An _is_hf_initialized flag regression is unrelated to CanineModel fp16 ONNX attention dtype mismatch." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both are loading failures, but one is a specific repo not loading and the other is a different transformers>=5.1.0 regression for another model." + }, + { + "left": "issue:36032", + "right": "issue:39921", + "accept": false, + "reason": "T5 tokenizer method conflict and Gemma3N special-token projection error are different tokenizer/model mechanisms." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "T5 tokenizer loading conflict and Qwen3Moe FineGrainedFP8Config failure are unrelated." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Different tokenizer load issues: repo-specific loading failure vs. Mistral regex patch crash." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Mistral-3 dtype mismatch in image preprocessing is unrelated to a template compilation TypeError." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Tokenizer BPE-marker handling and lm_head weight tying are different bugs in different model components." + }, + { + "left": "issue:43746", + "right": "issue:47331", + "accept": false, + "reason": "GraniteSpeech PEFT checkpoint loading and Metal quantized batched generate corruption are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:43644", + "accept": false, + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; not the same bug." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Mixtral grouped_mm tracing failure is unrelated to the non-template nodes compile error." + }, + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json is not the same as a DeepSeek-Coder tokenizer-output regression tied to a prior fix gap." + }, + { + "left": "issue:45479", + "right": "issue:46894", + "accept": false, + "reason": "Degenerate zero-loss classification bug and training-loss double-shift bug are different model/task errors." + } + ] + }, + "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": 28, + "review_unit_index": 11, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42915", + "issue:43208", + "issue:43262", + "issue:43299", + "issue:43317", + "issue:43421", + "issue:43452", + "issue:43482", + "issue:43582", + "issue:43632", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43792", + "issue:43873", + "issue:43931", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44530", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44987", + "issue:44991", + "issue:45084", + "issue:45479", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:46178", + "issue:46285", + "issue:46304", + "issue:46326", + "issue:46402", + "issue:46869", + "issue:46897", + "issue:47013", + "issue:47312", + "issue:47331" + ], + "soft_pairs": [ + "issue:44661|issue:44683", + "issue:46285|issue:46402", + "issue:43299|issue:43931", + "issue:42915|issue:44530", + "issue:45479|issue:46897", + "issue:43873|issue:47331", + "issue:44451|issue:44987", + "issue:44530|issue:46304", + "issue:36032|issue:43645", + "issue:44737|issue:45859", + "issue:40444|issue:46326", + "issue:40990|issue:43421", + "issue:34567|issue:40444", + "issue:42175|issue:46178", + "issue:45084|issue:45507", + "issue:43644|issue:43646", + "issue:43582|issue:43632", + "issue:44987|issue:44991", + "issue:39404|issue:44671", + "issue:44206|issue:44220", + "issue:43452|issue:43482", + "issue:43482|issue:47312", + "issue:43262|issue:47331", + "issue:43582|issue:47331", + "issue:44811|issue:45584", + "issue:43208|issue:47013", + "issue:42915|issue:45084", + "issue:43792|issue:46869", + "issue:43317|issue:43482" + ], + "prepared_review_unit_hash": "a66139a311f53acac9a40be05a5978338903526336d9fb12ba7d66651eb552ed", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36279, + "estimated_input_tokens": 9070, + "estimated_eval_tokens": 18396 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: many items are Transformers v5 regressions, but they concern different models, subsystems, and failure modes. Only the GGUF loading reports look like the same underlying bug; the rest should stay separate.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43646 is the broadest representative of the v5 regression bucket, but it is still only one specific custom-model-init failure and not a safe umbrella duplicate for the whole cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best representative issue because it is the most generic v5 breakage in the set, though it should not absorb the unrelated tokenizer/audio/quantization/offload reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Different code paths: tokenizer mapping/model registration vs compiled flex_attention on torch >= 2.9." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "Both are model-loading regressions, but for different models and different root causes." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Qwen loading failures, but one is FP8 config handling and the other is a weight-shape mismatch for a VL model." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Both are Qwen-related, but one is FineGrainedFP8Config loading and the other is a cache/group-type crash in PagedAttentionCache." + }, + { + "left": "issue:45479", + "right": "issue:46897", + "accept": false, + "reason": "Both affect training loss, but they describe different bugs in different models and should not be merged." + }, + { + "left": "issue:43873", + "right": "issue:47331", + "accept": false, + "reason": "Quantization/offload problems, but one is GPU offloading behavior and the other is a Metal batched-generate corruption bug." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Both are loading regressions, but they target different models and don't point to the same loader bug." + }, + { + "left": "issue:44530", + "right": "issue:46304", + "accept": false, + "reason": "Unrelated failures: Qwen3.5 cache handling vs a distributed-build/PyTorch support regression." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer method-name conflict vs custom model initialization in notebooks; not the same bug." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "Both touch device/offload behavior, but one is CPU placement in XLNet and the other is Qwen3_5Moe offload compatibility." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "Different subsystems and symptoms: multimodal dataset finetuning vs mistral_common chat-template behavior." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity quality issue vs tokenizer post-processor update behavior; not the same underlying bug." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "Trainer metrics tracking vs multimodal IterableDataset handling; unrelated." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "Both mention TensorFlow backends, but one is packaging extras and the other is a missing backend mapping entry." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Different failures: template compilation error vs an invalid method call in GraniteMoEHybrid." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 regressions, but one is non-persistent buffer initialization and the other is custom model initialization in notebooks." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "AppleSilicon allocator TypeError and _is_hf_initialized flag breakage are separate issues." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are model-loading regressions after v5, but they involve different models and likely different loader paths." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper pipeline regression vs CamemBERT masked-LM prediction bug; unrelated." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Same audio area, but the titles point to different feature-extraction failures without enough evidence of one root cause." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": true, + "reason": "Both point to the same GGUF loading regression in transformers v5; the Qwen2.5 case looks like a concrete instance of the broader gguf_file breakage." + }, + { + "left": "issue:43482", + "right": "issue:47312", + "accept": false, + "reason": "GGUF loading failure and kernel version incompatibility are unrelated." + }, + { + "left": "issue:43262", + "right": "issue:47331", + "accept": false, + "reason": "Audio chat-template sampling-rate bug vs Metal quantized batched generation corruption; different code paths." + }, + { + "left": "issue:43582", + "right": "issue:47331", + "accept": false, + "reason": "Allocator warmup TypeError on Apple Silicon is unrelated to the Metal quantization row-corruption bug." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both are Whisper-related, but one is skip_special_tokens handling in batch_decode and the other is empty-transcription failure after align_special_tokens." + }, + { + "left": "issue:43208", + "right": "issue:47013", + "accept": false, + "reason": "Both are xLSTM training bugs, but they describe distinct crashes and should be tracked separately." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Different subsystems and failures: Qwen3Moe FP8 loading vs template compilation." + }, + { + "left": "issue:43792", + "right": "issue:46869", + "accept": false, + "reason": "Whisper can't-run issue vs YOLO multi-GPU device-map failure; unrelated." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "dequantized device_map offload failure vs GGUF model loading failure; different problems." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative about the heterogeneous cluster. The only accepted soft edge, issue:43452 \u2194 issue:43482, is reasonably supported by the titles as the same GGUF-loading regression, while the other pairwise rejections avoid over-merging distinct model/subsystem bugs." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 12, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28530", + "issue:34567", + "issue:36032", + "issue:40444", + "issue:42915", + "issue:43089", + "issue:43262", + "issue:43296", + "issue:43322", + "issue:43377", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43540", + "issue:43632", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43746", + "issue:43819", + "issue:43872", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44849", + "issue:44877", + "issue:44987", + "issue:45020", + "issue:45081", + "issue:45325", + "issue:45923", + "issue:45987", + "issue:46326", + "issue:46661", + "issue:46821", + "issue:46899", + "issue:47254", + "issue:47448" + ], + "soft_pairs": [ + "issue:43746|issue:46326", + "issue:43645|issue:46661", + "issue:44403|issue:45081", + "issue:42915|issue:44291", + "issue:44987|issue:45081", + "issue:43646|issue:44534", + "issue:43262|issue:43377", + "issue:45020|issue:47448", + "issue:36032|issue:43742", + "issue:43377|issue:43819", + "issue:42915|issue:44589", + "issue:43994|issue:44610", + "issue:44008|issue:46899", + "issue:28530|issue:34567", + "issue:44062|issue:44987", + "issue:44451|issue:45081", + "issue:43531|issue:44877", + "issue:44265|issue:44610", + "issue:43404|issue:44466", + "issue:44849|issue:45325", + "issue:44479|issue:45325", + "issue:43632|issue:46821", + "issue:43322|issue:43540", + "issue:43089|issue:45987", + "issue:43425|issue:44403", + "issue:43296|issue:43531", + "issue:45923|issue:47254", + "issue:44464|issue:45987", + "issue:40444|issue:45325", + "issue:43872|issue:44291" + ], + "prepared_review_unit_hash": "206391d579b92f354bd1377993a3f836e9651aeb3751120c19ed54de8bf88587", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36965, + "estimated_input_tokens": 9242, + "estimated_eval_tokens": 18740 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are just thematically related transformer bugs, not true duplicates. The only clear duplicate-looking pair is the `_is_hf_initialized`/`init_empty_weights` regression affecting model loading, including the bitsandbytes case.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44291 is the clearest canonical report: it describes the underlying `init_empty_weights` / `_is_hf_initialized` breakage directly, while issue:43872 is a downstream bitsandbytes-specific manifestation of the same regression.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44291 is the best cluster representative because it names the shared root cause in a general way and is not tied to one downstream library or model variant.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43746", + "right": "issue:46326", + "accept": false, + "reason": "Different subsystems and failures: GraniteSpeech PEFT checkpoint loading vs Mistral-common tokenizer chat-template behavior." + }, + { + "left": "issue:43645", + "right": "issue:46661", + "accept": false, + "reason": "Both involve model initialization, but one is notebook/custom-model setup and the other is remote-code `config_class` lookup; not the same bug." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Unrelated symptoms: loading-log noise vs a Mistral tokenizer regex patch crash." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Both are model-loading failures, but one is Qwen3MoE FP8 config handling and the other is `_is_hf_initialized`/`init_empty_weights`; different code paths." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Different failures and model families; no clear shared underlying bug." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Custom model init regression vs non-persistent buffer serialization junk; too different to merge." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Both audio-related, but one is chat-template sampling-rate defaulting and the other is encoder padding-mask batching correctness." + }, + { + "left": "issue:45020", + "right": "issue:47448", + "accept": false, + "reason": "Generic remote-code breakage vs a specific model-architecture recognition problem; not the same concrete bug." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Tokenizer constructor conflict vs loading a specific model failing with KeyError; unrelated." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Different audio-model bugs: batching/padding-mask mismatch vs DAC latent/forward mismatch without STE." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Different failure modes and components; no shared code-path bug." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Both are vision-model issues, but one is wrong outputs from AutoModel/pipeline and the other is a processor size mismatch." + }, + { + "left": "issue:44008", + "right": "issue:46899", + "accept": false, + "reason": "Same general Gemma family, but the bugs differ: variable-name collision in forward vs unsafe dtype casting in multimodal embedders." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Training callback early-stopping logic vs token-count bookkeeping; unrelated." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "Token constructor argument conflict vs loading a specific model with recent transformers; unrelated." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "Both are loading failures, but for different models and different tokenizer internals." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior vs strict config rejection for granite_speech; unrelated." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export compatibility issue vs processor/model input-size mismatch; not the same bug." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both mention `lm_head.weight`, but one is missing weight tying in a model class and the other is device-dependent serialization; not the same code-path problem." + }, + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen2.5/Qwen3.5 behaviors: output_hidden_states bug vs temporal position-id scaling in Qwen2.5-VL." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Related Qwen VL area, but one is a broader video-input regression and the other is a still-image temporal position-id scaling bug." + }, + { + "left": "issue:43632", + "right": "issue:46821", + "accept": false, + "reason": "Both are API regressions, but one is `_is_hf_initialized` and the other is `window_size`; unrelated." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Different multimodal models and failures: Llava Next load crash vs Qwen3OmniMoe video-processing ValueError." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation performance overhead vs cache length return-type correctness; not the same bug." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch version compatibility vs loading-log noise; unrelated." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL/vLLM loading failure vs Qwen3-MoE sliding-window bug; no overlap." + }, + { + "left": "issue:45923", + "right": "issue:47254", + "accept": false, + "reason": "Both mention gradient checkpointing, but one is a missing flag on a remote-code variant and the other is a checkpointing/ZeRO-3 runtime error; different issues." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Chunked-generation inconsistency with compiled forward vs cache-length API contract bug; not the same concrete problem." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Qwen2.5-VL dataset/multi-image finetuning failure vs Qwen2.5-VL still-image temporal position IDs; different parts of the stack." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": true, + "reason": "Same underlying regression: transformers v5\u2019s `_is_hf_initialized`/`init_empty_weights` change breaks loading, and bitsandbytes is just a downstream manifestation." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is well grounded in the packet. The canonical representative choice of issue:44291 over issue:43872 is supported by the titles, and the one accepted soft pair is a clear shared regression around `_is_hf_initialized` / `init_empty_weights`. The remaining soft-edge rejections stay conservative and do not overclaim duplicates." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 13, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:33666", + "issue:35707", + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43296", + "issue:43381", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43611", + "issue:43632", + "issue:43716", + "issue:43746", + "issue:43792", + "issue:43883", + "issue:44079", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44387", + "issue:44403", + "issue:44488", + "issue:44534", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45216", + "issue:45357", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:46294", + "issue:46315", + "issue:46491", + "issue:46710", + "issue:46821", + "issue:46899", + "issue:47404", + "issue:47448" + ], + "soft_pairs": [ + "issue:44220|issue:44683", + "issue:43716|issue:46899", + "issue:43611|issue:44534", + "issue:44991|issue:46315", + "issue:44403|issue:44991", + "issue:44291|issue:44589", + "issue:45812|issue:46491", + "issue:45216|issue:45357", + "issue:42796|issue:43746", + "issue:43883|issue:44488", + "issue:43792|issue:44589", + "issue:33666|issue:40444", + "issue:43632|issue:44756", + "issue:43381|issue:45923", + "issue:35707|issue:40444", + "issue:40990|issue:44811", + "issue:43531|issue:44291", + "issue:44079|issue:45663", + "issue:42796|issue:46710", + "issue:46294|issue:47448", + "issue:41720|issue:45081", + "issue:43482|issue:43525", + "issue:43296|issue:46821", + "issue:44387|issue:47404", + "issue:42175|issue:43296", + "issue:43883|issue:44991", + "issue:42175|issue:45542", + "issue:44079|issue:44534", + "issue:43482|issue:45538", + "issue:44265|issue:45072" + ], + "prepared_review_unit_hash": "edc051a6b9094068099b584f217c4627e5d528052bfe8f184c6464362422aefb", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36771, + "estimated_input_tokens": 9193, + "estimated_eval_tokens": 18642 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly noisy similarity with unrelated issues, but two issue pairs look like true duplicates: Qwen3.5 save_pretrained key corruption (45216/45357) and TensorFlow backend packaging/install problems (42175/45542). No PRs are present.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 45216 is the best canonical anchor: it is the earliest, broadest report of the Qwen3.5 save_pretrained regression, and 45357 reads like a later follow-up of the same incorrect visual-encoder key serialization bug.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45216 is the strongest representative issue in the cluster because it names the concrete regression, has the clearest symptom description, and matches the later duplicate report most closely.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Both are torch-related failures, but they hit different code paths: fbank feature extraction vs compiled flex_attention on newer torch." + }, + { + "left": "issue:43716", + "right": "issue:46899", + "accept": false, + "reason": "Both mention dtype issues in multimodal models, but they are different models/components (Mistral-3 image preprocessing vs Gemma4 audio/vision embedders) with different failure modes." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is model loading with base_model_prefix and the other is non-persistent buffer initialization; not the same bug." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer-related, but one is a load failure for a specific tokenizer while the other is a performance regression in convert_tokens_to_ids." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "One is about noisy loading logs, the other is a tokenizer loading failure; different symptoms and fixes." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Both involve Transformers v5 compatibility, but they are distinct issues: unexpected _is_hf_initialized argument vs missing Float8 storage object." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and different behaviors (Granite token IDs vs CodeLlama leading-space round-trip)." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same underlying Qwen3.5 save_pretrained regression: both report incorrect checkpoint serialization, with 45357 acting as a later reproduction/follow-up of the same visual-encoder key bug." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Tokenizer BPE-marker handling and GraniteSpeech PEFT checkpoint loading are unrelated bugs." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Both are model loading failures, but one is a missing all_tied_weights_keys attribute and the other is a specific tokenizer/model load issue; not the same code path." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper run failure and Float8 storage lookup error are unrelated." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both are Qwen2-VL-family training issues, but one is multi-GPU training and the other is iterable-dataset finetuning with multiple images per prompt; different failures." + }, + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Transformers v5 _is_hf_initialized breakage and Strix Halo mmap OOM are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Both mention gradient checkpointing, but one is eval-mode usage and the other is a missing supports_gradient_checkpointing flag; distinct issues." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to multimodal finetuning with multiple images per prompt." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "GPT-OSS perplexity regression and Whisper batch_decode special-token handling are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights _is_hf_initialized argument are separate bugs." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment and Gemma4 FSDP2 shared_kv_states rebuild are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer BPE-marker handling and DeepSeek accuracy regression are different problems." + }, + { + "left": "issue:46294", + "right": "issue:47448", + "accept": false, + "reason": "Qwen3.5 video chat-template mismatch and Gemma4 architecture recognition are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "Qwen3 auto device mapping cudaErrorAssert and Mistral regex patch crashes are unrelated." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "GGUF loading failure in Transformers v5 and missing pad_token_id on Llama4Config are different incompatibilities." + }, + { + "left": "issue:43296", + "right": "issue:46821", + "accept": false, + "reason": "PaddleOCR-VL loading and TimesFM window_size AttributeError are unrelated." + }, + { + "left": "issue:44387", + "right": "issue:47404", + "accept": false, + "reason": "Both are OOM-related, but one is int4 quantization memory growth and the other is device_map transient allocation during conversion; different root causes." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "TensorFlow backend packaging and PaddleOCR-VL load failure are unrelated." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Missing all_tied_weights_keys and tokenizer loading for EMBEDDIA/est-roberta are unrelated." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": true, + "reason": "Both describe the same packaging/back-end dependency problem around TensorFlow being absent or incorrectly omitted while the tf backend is partially installed." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "ModelOutput key handling and non-persistent buffer junk initialization are unrelated." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "Qwen2.5-GGUF loading and CLIPTokenizer model_max_length behavior are unrelated." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "torch.export failures with torch_compilable_check and dtype mismatches in bfloat16 inference are different issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays appropriately conservative. The two accepted duplicate-like pairs are well supported by the issue titles and descriptions, and the rest of the soft-edge rejections avoid overclaiming same-bug matches. No PRs are present, as stated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 14, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:40444", + "issue:40990", + "issue:41720", + "issue:42796", + "issue:43317", + "issue:43377", + "issue:43421", + "issue:43454", + "issue:43606", + "issue:43611", + "issue:43644", + "issue:43646", + "issue:43818", + "issue:43828", + "issue:43856", + "issue:43873", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44462", + "issue:44466", + "issue:44479", + "issue:44509", + "issue:44568", + "issue:44661", + "issue:44829", + "issue:44877", + "issue:44898", + "issue:45325", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45491", + "issue:45584", + "issue:45701", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:46402", + "issue:46661", + "issue:46697", + "issue:46821", + "issue:46909", + "issue:47217", + "issue:47448" + ], + "soft_pairs": [ + "issue:44462|issue:45812", + "issue:46661|issue:47448", + "issue:43317|issue:43856", + "issue:45325|issue:45834", + "issue:43611|issue:43646", + "issue:40444|issue:45381", + "issue:44509|issue:46909", + "issue:44291|issue:44877", + "issue:43421|issue:44568", + "issue:43644|issue:44079", + "issue:43377|issue:45491", + "issue:45325|issue:45381", + "issue:43994|issue:44898", + "issue:40444|issue:47217", + "issue:43454|issue:44466", + "issue:44661|issue:47448", + "issue:43873|issue:45701", + "issue:45325|issue:46697", + "issue:43606|issue:44265", + "issue:43818|issue:45357", + "issue:43818|issue:46402", + "issue:44479|issue:45823", + "issue:42796|issue:45381", + "issue:43828|issue:46821", + "issue:44829|issue:45479", + "issue:41720|issue:45823", + "issue:40990|issue:45584", + "issue:44186|issue:44265" + ], + "prepared_review_unit_hash": "211454b1344a6bc4f3f113cc048e7817156c00a3dc1be83f8f6e2940eb277ccf", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35851, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose collection of mostly unrelated Transformers issues: tokenizer regressions, multimodal/video position-id bugs, device-map/offload failures, model-loading regressions, and docs problems. I don't see any soft pair that is clearly the same underlying bug or change, so none should be merged as duplicates. If a single issue must represent the cluster, issue #43873 is the most active and best-documented bug report, but it is not a duplicate anchor for the others.", + "confidence": 0.95, + "canonical_issue_reason": "#43873 is the strongest triage anchor: it is open, highly discussed, has inbound references, and describes a concrete reproducible offloading/quantization failure. It is still distinct from the other issues, so this is only a representative choice, not a duplicate match.", + "canonical_pr_reason": null, + "best_issue_reason": "#43873 is the most suitable issue to keep as the representative item because it has the most activity and the clearest end-user impact/repro among this otherwise heterogeneous set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both involve AutoTokenizer/tokenization regressions, but one is about ignoring tokenizer.json from the repo while the other is about wrong token IDs for Granite models; different failures and likely different code paths." + }, + { + "left": "issue:46661", + "right": "issue:47448", + "accept": false, + "reason": "Both concern loading/recognizing remote or new model architectures, but one is an AutoModel config_class AttributeError and the other is an unrecognized Gemma4 arch; distinct loader failures." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "One is a device_map/offload load failure with quantization, the other is a training-memory issue in Qwen3 MoE; same broad area, not the same bug." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Both are multimodal model bugs, but one is Qwen2.5-VL temporal position-id scaling and the other is a Kosmos2.5 OCR index error; different models and root causes." + }, + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both mention Transformers 5.0 loading regressions, but one is base_model_prefix-related loading and the other is custom model initialization; not the same concrete issue." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL multimodal regressions, but one is IterableDataset with multiple images per prompt and the other is video vision_position_ids; different input shapes and failure modes." + }, + { + "left": "issue:44509", + "right": "issue:46909", + "accept": false, + "reason": "Both are documentation issues, but one is about removed pipeline task docs and the other is a broken DIA example; unrelated doc problems." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "One is init_empty_weights/_is_hf_initialized loading incompatibility, the other is strict config rejecting granite_speech; separate regressions." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both involve special tokens, but one is runtime post-processor updates in TokenizersBackend and the other is add_special_tokens behavior for a specific tokenizer; not the same bug." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "Non-persistent buffer serialization junk and ModelOutput key assignment are different internal data-structure bugs." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "Both are batching/padding-related, but one is MIMI encoder missing padding-mask support and the other is Gemma3 NaNs from mixed-length sliding-window attention; different mechanisms." + }, + { + "left": "issue:45325", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL position-id issues, but one is still-image temporal scaling and the other is video-input vision_position_ids; related theme, different defect." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "One is a SigLIP2 AutoModel/pipeline output quality problem, the other is a Perceiver interpolate_pos_encoding failure; not the same bug." + }, + { + "left": "issue:40444", + "right": "issue:47217", + "accept": false, + "reason": "Both involve VL image placeholder handling, but one is finetuning with multiple images per prompt and the other is failure when the placeholder appears in user text; distinct scenarios." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both are tied-weight/lm_head-related, but one is AyaVision weight tying and the other is inconsistent lm_head serialization by device; different code paths." + }, + { + "left": "issue:44661", + "right": "issue:47448", + "accept": false, + "reason": "Both are model/arch recognition issues, but one is add-new-model-like failing inside TOKENIZER_MAPPING_NAMES and the other is Gemma4 arch detection; unrelated loader logic." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Quantization/offload behavior and tokenization changing across versions are unrelated regressions." + }, + { + "left": "issue:45325", + "right": "issue:46697", + "accept": false, + "reason": "Qwen2.5-VL position-id scaling and Swin2SR over-padding are unrelated preprocessing/model bugs." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch for bark-small and torch.export failures with torch_compilable_check are separate backend issues." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA temporal-attention/weight-sharing bug and Qwen3.5 save_pretrained visual-key regression are different model-specific problems." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Video-LLaVA architecture/weights issue and Sam3Model text encoder weight loading are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Both are device/version regressions in multimodal models, but one is Qwen2.5-VL video input and the other is Gemma4 PLE device mismatch; not the same concrete failure." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video vision_position_ids are unrelated bugs." + }, + { + "left": "issue:43828", + "right": "issue:46821", + "accept": false, + "reason": "Autocast dtype mismatch in Phi-tiny-MoE and TimesFM window_size AttributeError are completely different issues." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both describe degenerate training outcomes, but one is tied to flash_attention_3 and the other to single_label_classification with num_labels=1; separate causes." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Both are device_map='auto' failures, but one is a cudaErrorAssert on A800 for Qwen3 and the other is a device mismatch in Gemma4 PLE; distinct bugs." + }, + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "High perplexity on gpt-oss-20b and Whisper empty-transcription failure after align_special_tokens are unrelated model-specific issues." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER/padding crashes and torch.export failures with torch_compilable_check 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": 28, + "review_unit_index": 15, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:34634", + "issue:35707", + "issue:36331", + "issue:39921", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42796", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43262", + "issue:43317", + "issue:43322", + "issue:43377", + "issue:43454", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43646", + "issue:43720", + "issue:43746", + "issue:43749", + "issue:43819", + "issue:43873", + "issue:44387", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45325", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45920", + "issue:45923", + "issue:46050", + "issue:46209", + "issue:46315", + "issue:46821", + "issue:47404", + "issue:47405", + "issue:47458" + ], + "soft_pairs": [ + "issue:34634|issue:35707", + "issue:35707|issue:41720", + "issue:43262|issue:43746", + "issue:43646|issue:47458", + "issue:43606|issue:43819", + "issue:45823|issue:46821", + "issue:42617|issue:42915", + "issue:43454|issue:43550", + "issue:40990|issue:42947", + "issue:43257|issue:45923", + "issue:43454|issue:45072", + "issue:47404|issue:47405", + "issue:44568|issue:45920", + "issue:43377|issue:44610", + "issue:42796|issue:46209", + "issue:36331|issue:42617", + "issue:36331|issue:39921", + "issue:36331|issue:42915", + "issue:41720|issue:43322", + "issue:44877|issue:44991", + "issue:43749|issue:44534", + "issue:43454|issue:43720", + "issue:43317|issue:43873", + "issue:43377|issue:44898", + "issue:41720|issue:42617", + "issue:41720|issue:44977", + "issue:45325|issue:46315", + "issue:44387|issue:45704", + "issue:40990|issue:45701", + "issue:43322|issue:46050", + "issue:43572|issue:45759" + ], + "prepared_review_unit_hash": "c013198d83dada1b4701d94bab6ec91d0eae2e44c1db06ad27badac470a462b1", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37116, + "estimated_input_tokens": 9279, + "estimated_eval_tokens": 18814 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated bug reports spanning tokenization, offloading, quantization, configs, and model-specific failures. I do not see any soft pair that is clearly the same underlying issue, so I reject all soft edges. There are no PRs in the set; if a single issue must anchor the cluster, the broadest/highest-activity report is 43873.", + "confidence": 0.44, + "canonical_issue_reason": "Issue 43873 is the broadest and most active report here (offloading/quantization, open, multiple inbound refs), so it is the best cluster anchor even though the set is not a true duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "43873 has the widest apparent scope and strongest discussion/ref signal, making it the most plausible representative issue among otherwise distinct reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "Different subsystems and failure modes: BarkProcessor voice preset vs generation with inputs_embeds/past_key_values." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Unrelated generation-cache bug vs Qwen3 device-mapping cuda assert." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sampling-rate default vs GraniteSpeech PEFT checkpoint loading; different code paths." + }, + { + "left": "issue:43646", + "right": "issue:47458", + "accept": false, + "reason": "General Transformers 5.0 custom init breakage vs GPT-2 c_proj initialization bug; not the same defect." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Both involve model behavior, but one is Bark CPU offload device mismatch and the other is DAC latents/STE mismatch." + }, + { + "left": "issue:45823", + "right": "issue:46821", + "accept": false, + "reason": "Gemma4 device-mapped forward mismatch vs TimesFM window_size AttributeError; distinct components." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3d_parallel.py runtime failure vs Qwen3Moe FP8 config failure; no shared concrete bug." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "AyaVision lm_head tying bug vs Bamba torch.compile/SDPA issue; different model and root cause." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "Perplexity regression on gpt-oss vs gradient checkpointing ineffective with LoRA; unrelated symptoms." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "MOE weight conversion with accelerate+deepspeed vs missing supports_gradient_checkpointing flag; separate issues." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "AyaVision weight-tying bug vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:47404", + "right": "issue:47405", + "accept": false, + "reason": "Both are loading-related, but one is OOM during conversion transients and the other is silent wrong-Auto-class conversion skip." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both are tokenizer regressions, but one is missing BOS/EOS for a specific tokenizer and the other is wrong token IDs across several models." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "MIMI padding-mask batching bug vs OmDet-Turbo processor image-size mismatch; different modalities and fixes." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE-marker handling vs FP8 inference failure on B200; no shared code path." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer loss signature regression vs parallel script execution failure; unrelated." + }, + { + "left": "issue:36331", + "right": "issue:39921", + "accept": false, + "reason": "Unexpected trainer kwarg vs Gemma3N special-token projection error; different layers of the stack." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "Trainer API mismatch vs Qwen3Moe FP8 config failure; no overlap beyond being runtime errors." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 device-map cuda assert vs Llava Next segmentation fault; different models and failure mechanisms." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config loading error vs tokenizer loading regression; not the same bug." + }, + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "FSDP CPU RAM efficient loading breakage vs non-persistent buffers being filled with junk; both v5 regressions but distinct." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "AyaVision weight tying issue vs BitNet packed-weight unpacking during accelerate load; separate model-loading paths." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both touch offloading, but one is dequantized-model loading on gpu+cpu offload and the other is broad quantized offloading behavior; too different to merge." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "MIMI encoder padding-mask mismatch vs Perceiver non-default resolution interpolation failure; unrelated architecture and bug." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Device-mapping assert on Qwen3 vs failing 3d_parallel.py; no common underlying defect." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Same model family only at a high level; flash-attention generation issue is not the same as auto device-map cuda assert." + }, + { + "left": "issue:45325", + "right": "issue:46315", + "accept": false, + "reason": "Qwen2.5-VL rope/position-id scaling bug vs slow convert_tokens_to_ids regression; different tokenizer/multimodal code paths." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Int4 quantization memory regression vs apex FusedRMSNorm memory leak; both memory-related but not the same bug." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "High perplexity regression vs version-dependent tokenization change; not enough to treat as one defect." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Segfault loading Llava Next vs quantized model save/load breakage; different symptoms and models." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig vs unwrap text_config for composite Qwen multimodal configs; unrelated 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": 28, + "review_unit_index": 16, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:26413", + "issue:28530", + "issue:33666", + "issue:41553", + "issue:41720", + "issue:42915", + "issue:43065", + "issue:43295", + "issue:43317", + "issue:43404", + "issue:43450", + "issue:43531", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43646", + "issue:43708", + "issue:43720", + "issue:43873", + "issue:44008", + "issue:44186", + "issue:44423", + "issue:44451", + "issue:44466", + "issue:44610", + "issue:44756", + "issue:44898", + "issue:45072", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45685", + "issue:45701", + "issue:45704", + "issue:45823", + "issue:46291", + "issue:46402", + "issue:46612", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:46869", + "issue:46897", + "issue:47020", + "issue:47317", + "issue:47404" + ], + "soft_pairs": [ + "issue:43450|issue:45701", + "issue:41553|issue:43582", + "issue:45685|issue:46821", + "issue:43873|issue:44756", + "issue:44186|issue:45072", + "issue:44466|issue:45357", + "issue:43295|issue:45701", + "issue:26413|issue:28530", + "issue:44451|issue:47020", + "issue:43065|issue:46402", + "issue:43550|issue:43606", + "issue:43550|issue:43720", + "issue:43577|issue:43708", + "issue:46821|issue:46869", + "issue:44186|issue:44610", + "issue:46612|issue:46858", + "issue:44008|issue:45823", + "issue:41720|issue:45161", + "issue:44423|issue:45406", + "issue:43646|issue:46665", + "issue:42915|issue:43317", + "issue:46897|issue:47317", + "issue:44610|issue:44898", + "issue:42915|issue:46291", + "issue:28530|issue:33666", + "issue:43404|issue:45357", + "issue:45704|issue:47404", + "issue:43531|issue:43572" + ], + "prepared_review_unit_hash": "ffc211e0c6f763fb8feb9b1d991149732cf3039e3084be094cccfb913b1204cc", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36226, + "estimated_input_tokens": 9057, + "estimated_eval_tokens": 18370 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose topical grouping, not a duplicate set: the items cover many unrelated bugs across trainer resume logic, tokenizers/processors, multimodal serving, device mapping/offload, quantization, and serialization. None of the proposed soft edges look like the same concrete underlying bug/change, so I would not merge any of them.", + "confidence": 0.91, + "canonical_issue_reason": "No single canonical duplicate stands out; the items span distinct bugs and subsystems rather than one shared underlying issue.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43708 is the strongest standalone report: it is specific, reproducible, and describes a concrete Trainer regression in checkpoint resume/max_steps calculation.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Different problems: batched video processor output shape vs version-dependent tokenization behavior." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "Unrelated: bad AutoTokenizer error for Voxtral vs Apple Silicon TypeError in caching_allocator_warmup." + }, + { + "left": "issue:45685", + "right": "issue:46821", + "accept": false, + "reason": "Different components and failures: MPS MoE histogram issue vs TimesFM window_size AttributeError." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Both involve memory/offload, but the concrete bugs differ: quantization offload behavior vs mmap OOM on Strix Halo." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash on NER/batched padding is unrelated to bfloat16 dtype mismatches in model inference." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both are serialization-related, but one is lm_head weight tying and the other is visual encoder key saving." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "Regression around processor.tokenizer/images is not the same as generic tokenization changes across versions." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Different Trainer issues: resume_from_checkpoint epoch_iterator failure vs early stopping metric lookup/disablement." + }, + { + "left": "issue:44451", + "right": "issue:47020", + "accept": false, + "reason": "Different tokenizer loading bugs for different models and code paths." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "Different SAM3 bugs: dummy Conv2d in pixel decoder vs text encoder weights not loading." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Different model/runtime failures: Bamba torch.compile+SDPA vs Bark CPU offload device mismatch." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "Different concrete causes: SDPA/compile failure vs packed weights not unpacked during accelerate loading." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Blip2 dtype propagation and Trainer checkpoint-resume max_steps are unrelated bugs." + }, + { + "left": "issue:46821", + "right": "issue:46869", + "accept": false, + "reason": "TimesFM window_size AttributeError is unrelated to YOLO multi-GPU device_map auto loading." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "Different vision/text preprocessing bugs with different root causes and affected models." + }, + { + "left": "issue:46612", + "right": "issue:46858", + "accept": false, + "reason": "Both concern generation/cache internals, but one is missing beam cache reorder support and the other is a device-tensor arithmetic issue in static cache prefill." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Different multimodal model bugs: Gemma3n forward variable collision vs Gemma4 PLE device mismatch." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Distinct parallelism/device-placement failures in different models." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both involve serve/processor paths, but one is a string-to() crash in continuous batching and the other is a missing _tokenizer attribute." + }, + { + "left": "issue:43646", + "right": "issue:46665", + "accept": false, + "reason": "Related to model loading/init, but not the same concrete bug or fix." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Different loading/offload failures: FineGrainedFP8Config on Qwen3Moe vs dequantized model loading with gpu+cpu offload." + }, + { + "left": "issue:46897", + "right": "issue:47317", + "accept": false, + "reason": "Same broad symptom category, but different models and likely different loss-shift causes." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Different image-size preprocessing bugs in different model families." + }, + { + "left": "issue:42915", + "right": "issue:46291", + "accept": false, + "reason": "FP8 model loading and kernels package version/revision import failure are unrelated." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early stopping behavior is unrelated to Qwen2-VL multi-GPU training." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Different weight-serialization problems affecting different model components." + }, + { + "left": "issue:45704", + "right": "issue:47404", + "accept": false, + "reason": "Both are memory-related, but the causes differ: apex fused RMSNorm leak vs unbudgeted contiguous() OOM during device_map loading." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Different issues: Qwen3-MoE sliding_window behavior vs StableLmConfig pad_token_idx regression." + } + ] + }, + "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": 28, + "review_unit_index": 17, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28530", + "issue:29159", + "issue:30064", + "issue:33357", + "issue:34634", + "issue:35707", + "issue:36032", + "issue:41720", + "issue:42915", + "issue:43257", + "issue:43316", + "issue:43322", + "issue:43441", + "issue:43452", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43653", + "issue:43716", + "issue:43746", + "issue:43782", + "issue:43873", + "issue:43881", + "issue:44186", + "issue:44361", + "issue:44387", + "issue:44464", + "issue:44488", + "issue:44534", + "issue:44671", + "issue:44734", + "issue:44743", + "issue:44913", + "issue:45479", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:46894", + "issue:47020", + "issue:47254", + "issue:47317", + "issue:47337" + ], + "soft_pairs": [ + "issue:43716|issue:46821", + "issue:28530|issue:30064", + "issue:43441|issue:47337", + "issue:43881|issue:44488", + "issue:44186|issue:44361", + "issue:44464|issue:46858", + "issue:44734|issue:46858", + "issue:30064|issue:34634", + "issue:29159|issue:36032", + "issue:43452|issue:47020", + "issue:42915|issue:43322", + "issue:44671|issue:44743", + "issue:44387|issue:45701", + "issue:43873|issue:45704", + "issue:43531|issue:45759", + "issue:33357|issue:35707", + "issue:45479|issue:47317", + "issue:43540|issue:43782", + "issue:43316|issue:44913", + "issue:43550|issue:43653", + "issue:45632|issue:45684", + "issue:44534|issue:46665", + "issue:41720|issue:43746", + "issue:46894|issue:47317", + "issue:43257|issue:47254", + "issue:34634|issue:41720" + ], + "prepared_review_unit_hash": "0ac861be93c45f4c824fea12ed2e172a078d8d6d85014a2755c0245f53dc316f", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34793, + "estimated_input_tokens": 8699, + "estimated_eval_tokens": 17654 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This looks like a loose similarity cluster of mostly unrelated Transformers issues rather than true duplicates. The soft pairs share broad subsystem language (tokenizers, cache, generation, multimodal loading, quantization), but the concrete failure modes and code paths differ, so none should be merged as duplicates.", + "confidence": 0.21, + "canonical_issue_reason": "issue:43873 has the strongest triage signal in this set: it is active, has the most inbound references and discussion, and describes a concrete reproducible bug with broad user impact.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the most representative issue to anchor the cluster because it is well-discussed, externally referenced, and clearly scoped as a real bug report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43716", + "right": "issue:46821", + "accept": false, + "reason": "Different products and failures: image preprocessor dtype mismatch vs a window_size AttributeError. Not the same underlying bug." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Early stopping/eval metric handling and void segmentation maps are unrelated code paths." + }, + { + "left": "issue:43441", + "right": "issue:47337", + "accept": false, + "reason": "FlashAttention failure for a specific model/version is unrelated to a flaky bbox range test failure." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading issues, but they concern different models and likely different loader bugs; too broad to merge." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Tokenizer crashes, but on different tokenizer classes and different trigger conditions; not the same bug." + }, + { + "left": "issue:44464", + "right": "issue:46858", + "accept": false, + "reason": "Both involve generation/cache/torch.compile, but one is inconsistent chunked generation output and the other is a static-cache length tensor arithmetic bug on non-CUDA backends." + }, + { + "left": "issue:44734", + "right": "issue:46858", + "accept": false, + "reason": "Serving KV-cache continuation indexing bug is distinct from the device-tensor cache length arithmetic issue." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "Image segmentation-map processing and BarkProcessor voice_preset behavior are unrelated." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Generic slow/fast tokenizer inconsistency is not the same as T5 tokenizer class-name conflict on load." + }, + { + "left": "issue:43452", + "right": "issue:47020", + "accept": false, + "reason": "GGUF/autoloading breakage and an XLMRoberta Unigram constructor mismatch are different tokenizer/model-init bugs." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Llava Next loading segfault are different model-loading paths and failures." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "Masked LM prediction regression in CamemBERT is unrelated to Qwen recurrent-state reset behavior." + }, + { + "left": "issue:44387", + "right": "issue:45701", + "accept": false, + "reason": "Quantization memory/OOM regression is not the same as a tokenization-version-change issue." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantization offloading behavior and an apex FusedRMSNorm memory leak are distinct bugs." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and missing text_config unwrapping in composite multimodal configs are different implementation problems." + }, + { + "left": "issue:33357", + "right": "issue:35707", + "accept": false, + "reason": "MacOS bus error on a CLIP model and progressive generation with inputs_embeds/past_key_values are unrelated." + }, + { + "left": "issue:45479", + "right": "issue:47317", + "accept": false, + "reason": "Both are training-loss bugs, but one is a general single-label classification edge case and the other is PPFormulaNet-specific." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Video input processing in Qwen3OmniMoe and Qwen3VL weight_only loading errors are different multimodal bugs." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "API discrepancy in Gemma3TextConfig is unrelated to GPTNeoX rotary_pct reload persistence." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba and BigBirdTokenizer special-token registration are unrelated." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Cache-path collisions for trust_remote_code and read-only permission propagation on save_pretrained are different filesystem issues." + }, + { + "left": "issue:44534", + "right": "issue:46665", + "accept": false, + "reason": "Non-persistent buffer initialization junk and from_pretrained-in-Python-class behavior are not the same code-path problem." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3 device-mapping cudaErrorAssert and GraniteSpeech PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:46894", + "right": "issue:47317", + "accept": false, + "reason": "Both mention training loss, but the models and underlying loss computation bugs differ; same class of symptom is not enough." + }, + { + "left": "issue:43257", + "right": "issue:47254", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed and a PEFT/ZeRO-3 checkpoint error are different distributed-loading issues." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice_preset and Qwen3 auto device-mapping cuda assert 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": 28, + "review_unit_index": 18, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29159", + "issue:29925", + "issue:30064", + "issue:34567", + "issue:35707", + "issue:36010", + "issue:39692", + "issue:42490", + "issue:43329", + "issue:43388", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43606", + "issue:43644", + "issue:43653", + "issue:43720", + "issue:43881", + "issue:43883", + "issue:44062", + "issue:44077", + "issue:44451", + "issue:44492", + "issue:44509", + "issue:44589", + "issue:44610", + "issue:44625", + "issue:44898", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45059", + "issue:45072", + "issue:45357", + "issue:45406", + "issue:45464", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:45923", + "issue:46402", + "issue:46665", + "issue:46821", + "issue:46858", + "issue:47333", + "issue:47405", + "issue:47448" + ], + "soft_pairs": [ + "issue:45406|issue:45464", + "issue:43606|issue:43720", + "issue:43531|issue:46821", + "issue:42490|issue:43329", + "issue:43531|issue:44589", + "issue:29159|issue:29925", + "issue:34567|issue:35707", + "issue:43881|issue:44991", + "issue:44492|issue:44509", + "issue:45759|issue:45859", + "issue:43606|issue:43653", + "issue:43540|issue:46858", + "issue:44451|issue:44991", + "issue:45704|issue:47405", + "issue:43388|issue:45059", + "issue:43526|issue:43883", + "issue:44987|issue:45042", + "issue:45042|issue:47448", + "issue:45923|issue:47333", + "issue:44077|issue:44625", + "issue:43644|issue:46665", + "issue:44898|issue:45072", + "issue:45357|issue:46402", + "issue:39692|issue:44610", + "issue:44062|issue:44451", + "issue:30064|issue:36010" + ], + "prepared_review_unit_hash": "07c20916d79d5b13e0f28a60fc71cf94a741eecf893ff0872ece51aff99fa754", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35071, + "estimated_input_tokens": 8768, + "estimated_eval_tokens": 17792 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a mixed bag of unrelated Transformers issues; every soft pair looks like thematic similarity rather than the same underlying bug/change, so no duplicates are accepted.", + "confidence": 0.93, + "canonical_issue_reason": "No single canonical issue emerges; the items span unrelated tokenizer, loading, multimodal, training, and docs problems.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue; the set is too heterogeneous to represent one bug/change.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing `_tokenizer` in serve vs Qwen3.5 streaming chat/completions; different failure paths and components." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve loading/offload, but one is a CPU offload device mismatch and the other is BitNet packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43531", + "right": "issue:46821", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and TimesFM `window_size` AttributeError are unrelated model-specific bugs." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Trailing-slash local model import warning is unrelated to a multimodal video-token counting bug." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Sliding window handling in Qwen3-MoE is unrelated to the Float8 storage lookup error." + }, + { + "left": "issue:29159", + "right": "issue:29925", + "accept": false, + "reason": "General slow/fast tokenizer inconsistency is not the same as SigLIP-specific enforcement of `use_fast=True`." + }, + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "Trainer token-count tracking and progressive generation with `inputs_embeds`/`past_key_values` are different code paths." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "GLM-4V model loading failure and est-roberta tokenizer loading failure are distinct model-specific issues." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A cache-strategy typo and stale docs about removed pipelines are unrelated documentation issues." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Unwrapping `text_config` for composite Qwen multimodal configs is unrelated to Qwen3_5Moe rotary embedding CPU-offload incompatibility." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to BigBirdTokenizer mask token special-token registration." + }, + { + "left": "issue:43540", + "right": "issue:46858", + "accept": false, + "reason": "Qwen3OmniMoe video-input validation bug and static-cache seq-length tensor arithmetic bug are separate failures." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "ScandiBERT load failure and est-roberta tokenizer load failure are different checkpoints and symptoms." + }, + { + "left": "issue:45704", + "right": "issue:47405", + "accept": false, + "reason": "T5 apex RMSNorm memory leak is unrelated to silent model_type mismatch conversion skipping." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "`gather_for_metrics` tuple-label truncation and SAM3 text/box odd behavior are different training/inference problems." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to Molmo missing `all_tied_weights_keys`." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "A model-loading failure for physical-intelligence/fast is not the same as PIL processors incorrectly requiring torchvision." + }, + { + "left": "issue:45042", + "right": "issue:47448", + "accept": false, + "reason": "Torchvision dependency regression in PIL processors is unrelated to Gemma4 architecture recognition." + }, + { + "left": "issue:45923", + "right": "issue:47333", + "accept": false, + "reason": "Missing `supports_gradient_checkpointing` and disk-offload save failure are different model lifecycle bugs." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "patchtsmixer post-init policy and Qwen3.5 label propagation concern unrelated configs." + }, + { + "left": "issue:43644", + "right": "issue:46665", + "accept": false, + "reason": "Non-persistent buffer initialization junk and `from_pretrained` inside Python classes are unrelated behaviors." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Perceiver non-default-resolution failure and bfloat16 dtype mismatches are separate model/backend issues." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Qwen3.5 `save_pretrained` key regression is unrelated to Sam3Model text encoder weights not loading." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "SigLIP2 documentation example errors are unrelated to OmDet-Turbo processor/model resolution mismatch." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "AddedToken keyword duplication and ScandiBERT loading failure are unrelated tokenizer/model issues." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Void segmentation-map processing and `GenerationMixin` import error have no shared underlying code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 19, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:39692", + "issue:42886", + "issue:43381", + "issue:43421", + "issue:43540", + "issue:43596", + "issue:43673", + "issue:43720", + "issue:43746", + "issue:43782", + "issue:43819", + "issue:43844", + "issue:43873", + "issue:43881", + "issue:43927", + "issue:43931", + "issue:43994", + "issue:44077", + "issue:44220", + "issue:44368", + "issue:44464", + "issue:44514", + "issue:44545", + "issue:44625", + "issue:44671", + "issue:44843", + "issue:44857", + "issue:44918", + "issue:44936", + "issue:44995", + "issue:45081", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45479", + "issue:45907", + "issue:46032", + "issue:46050", + "issue:46294", + "issue:46463", + "issue:46655", + "issue:46693", + "issue:46858", + "issue:47050", + "issue:47337", + "issue:47448", + "issue:47458" + ], + "soft_pairs": [ + "issue:44918|issue:44936", + "issue:44514|issue:44545", + "issue:43746|issue:46050", + "issue:42886|issue:44843", + "issue:44625|issue:45479", + "issue:43746|issue:43873", + "issue:46655|issue:46693", + "issue:44464|issue:47050", + "issue:44995|issue:46858", + "issue:44077|issue:47458", + "issue:39692|issue:43873", + "issue:43381|issue:43844", + "issue:45081|issue:45356", + "issue:43782|issue:45381", + "issue:43673|issue:47337", + "issue:44857|issue:47050", + "issue:46032|issue:46463", + "issue:43931|issue:47458", + "issue:43881|issue:47448", + "issue:44368|issue:45357", + "issue:43994|issue:44671", + "issue:39692|issue:47050", + "issue:43421|issue:43927", + "issue:45356|issue:45907", + "issue:43540|issue:46294", + "issue:43720|issue:43819", + "issue:43596|issue:44220" + ], + "prepared_review_unit_hash": "a39f79417584fcbea8573271720cf1b1ef4b07b3725b44f543fb1204059f305c", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35543, + "estimated_input_tokens": 8886, + "estimated_eval_tokens": 18028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated bug reports that only weakly resemble each other by broad themes like tokenizers, model loading, caching, or multimodal preprocessing. Only two soft pairs look like true duplicates or near-duplicates: the Qwen2.5-VL batched padding crash pair, and the HF_HUB_OFFLINE tokenizer regression pair.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue cleanly covers the whole cluster, but issue 44843 is the best anchor because it identifies a concrete offline-tokenizer regression and its likely root cause in _patch_mistral_regex/model_info. The exact-title duplicate pair 44514/44545 is narrower and less representative of the cluster as a whole.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44843 is the strongest standalone candidate: it is specific, reproducible, and points to a clear code path, and it aligns with one of the accepted soft duplicates. Issue 44545 is an exact duplicate of 44514, but it is narrower in scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "Both involve training/evaluation behavior, but they describe different failures and likely different code paths." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same bug: identical Qwen2_5_VLProcessor.apply_chat_template crash on batched input with padding=False." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "Both are save/load-related, but one is PEFT local checkpoint loading and the other is post-save quantized model corruption." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": true, + "reason": "Same underlying offline tokenizer regression; 44843 pins the root cause to _patch_mistral_regex/model_info." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "Both concern classification configs, but one is num_labels propagation and the other is a degenerate loss formulation." + }, + { + "left": "issue:43746", + "right": "issue:43873", + "accept": false, + "reason": "Local checkpoint loading with adapters is not the same as quantization/offloading behavior." + }, + { + "left": "issue:46655", + "right": "issue:46693", + "accept": false, + "reason": "Video padding calculation and FA2 long-context performance regression are different problems." + }, + { + "left": "issue:44464", + "right": "issue:47050", + "accept": false, + "reason": "Chunked generation inconsistency under compiled forward is unrelated to ONNX fp16 attention dtype mismatch." + }, + { + "left": "issue:44995", + "right": "issue:46858", + "accept": false, + "reason": "Both mention cache, but one is a stale indexer cache crash and the other is a static-cache tensor-type arithmetic issue." + }, + { + "left": "issue:44077", + "right": "issue:47458", + "accept": false, + "reason": "Post-init allowance in patchtsmixer is unrelated to GPT2 c_proj initialization." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "Documentation/example errors are unrelated to quantization/offloading behavior." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Gradient checkpointing in eval mode and gradient growth under ZeRO-3 are different issues." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both touch tokenizer regex handling, but one is a crash and the other is a regression around codec handling/warnings." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen/VL problems with different symptoms and code paths." + }, + { + "left": "issue:43673", + "right": "issue:47337", + "accept": false, + "reason": "Generation cache behavior and a bbox-range flaky test are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:47050", + "accept": false, + "reason": "Different models and different failure modes; only a superficial fp16/CUDA theme is shared." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Cache correctness in Mamba2Mixer is unrelated to a T5 error-message bug." + }, + { + "left": "issue:43931", + "right": "issue:47458", + "accept": false, + "reason": "Model shape mismatch during load is unrelated to GPT2 weight initialization." + }, + { + "left": "issue:43881", + "right": "issue:47448", + "accept": false, + "reason": "Both are model-loading problems, but with different models and different causes." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "A warning about tie_word_embeddings is unrelated to incorrect visual encoder keys on save_pretrained." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Both affect model outputs, but SigLIP2 pipeline issues and CamemBERT MLM regression are distinct." + }, + { + "left": "issue:39692", + "right": "issue:47050", + "accept": false, + "reason": "Documentation example problems do not match an ONNX export dtype mismatch." + }, + { + "left": "issue:43421", + "right": "issue:43927", + "accept": false, + "reason": "Runtime post-processor updates and DiaConfig token-ID loss are separate tokenizer/config bugs." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer codec/regression handling is unrelated to list_repo_files revision forwarding." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-input failures, but they involve different models and different failure mechanisms." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "BitNet packed-weight loading and DAC latent conversion/STE mismatch are unrelated." + }, + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "DeepSpeed/BertModel zero3 init crash is unrelated to fbank feature extraction." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet and stays conservative. The two accepted soft pairs are well-supported by the issue titles and descriptions, and the rejected pairs are framed cautiously rather than overclaimed. No obvious evidence inflation 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": 28, + "review_unit_index": 20, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:30064", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:39921", + "issue:41720", + "issue:43295", + "issue:43421", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43611", + "issue:43643", + "issue:43653", + "issue:43720", + "issue:43819", + "issue:43927", + "issue:44008", + "issue:44062", + "issue:44186", + "issue:44220", + "issue:44488", + "issue:44561", + "issue:44661", + "issue:44821", + "issue:44843", + "issue:44863", + "issue:44898", + "issue:44936", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45381", + "issue:45834", + "issue:45987", + "issue:46650", + "issue:46665", + "issue:46693", + "issue:46829", + "issue:46832", + "issue:46858", + "issue:47020", + "issue:47050", + "issue:47405", + "issue:47458" + ], + "soft_pairs": [ + "issue:44821|issue:45084", + "issue:43550|issue:43819", + "issue:43550|issue:43927", + "issue:43421|issue:44186", + "issue:43421|issue:43653", + "issue:43295|issue:45834", + "issue:45072|issue:47050", + "issue:30064|issue:36331", + "issue:43540|issue:44008", + "issue:44488|issue:47020", + "issue:43653|issue:43720", + "issue:44062|issue:44488", + "issue:45381|issue:46693", + "issue:41720|issue:44991", + "issue:39921|issue:44936", + "issue:43531|issue:43643", + "issue:44220|issue:44661", + "issue:43611|issue:46665", + "issue:46829|issue:46832", + "issue:44008|issue:46858", + "issue:34634|issue:36010", + "issue:44186|issue:44898", + "issue:43653|issue:43819", + "issue:44863|issue:47458", + "issue:46650|issue:47405", + "issue:44991|issue:47020", + "issue:44843|issue:45081", + "issue:44561|issue:45020", + "issue:39692|issue:45834", + "issue:45987|issue:46829" + ], + "prepared_review_unit_hash": "2da3904f7c13b22b4da22b60e221c3d62b9eb692160bae1439b15393936fac17", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37063, + "estimated_input_tokens": 9266, + "estimated_eval_tokens": 18788 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is highly heterogeneous: the soft pairs are mostly same-subsystem but not same-bug matches, so this cluster does not support duplicate merging. I found one near-adjacent tokenizer-regex pair, but even that looks like two distinct failures in the same helper rather than one duplicate.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items span unrelated bugs across tokenizers, vision, training, model loading, and cache behavior, so there is no single duplicate family to anchor on.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a strong global representative because the approximate matches are mostly false positives; the closest items still describe different concrete failure modes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "Both are image/tokenizer-related only at a very broad level; one is AutoImageProcessor URL loading, the other is a compile/template-node error. Different code paths and fixes." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Different model families and failures: SDPA/torch.compile on Bamba vs missing STE in DAC.from_latents. Not the same underlying bug." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "One is a compile/SDPA failure, the other is token ID loss after save/load in DiaConfig. Unrelated root causes." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "Both involve tokenizers, but one is runtime post-processor updates and the other is LayoutLMv2 NER/padding crashes. Different behavior and likely different fixes." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Post-processor update bug vs BigBird mask token registration/empty decode. Same area, but not the same concrete defect." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor.tokenizer regression vs Kosmos2.5 long-OCR index error. Different models and failure mechanisms." + }, + { + "left": "issue:45072", + "right": "issue:47050", + "accept": false, + "reason": "Both mention dtype mismatches, but in different models and pipelines (SwitchTransformers/TimmWrapperModel inference vs CanineModel ONNX export). Not mergeable as one fix." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Void segmentation map processing vs CustomTrainer.compute_loss signature break. Completely unrelated." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Video-input ValueError in Qwen3OmniMoe vs Gemma3n variable-name collision in forward(). Different models and bugs." + }, + { + "left": "issue:44488", + "right": "issue:47020", + "accept": false, + "reason": "Both are tokenizer-loading problems, but one is a model-specific load failure and the other is an XLMRobertaTokenizer constructor type mismatch. Not the same issue." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "BigBird special-token decoding bug vs BitNet packed-weight unpacking during accelerate loading. Different subsystems and fixes." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken special-argument duplication vs loading a specific tokenizer model. No shared concrete failure." + }, + { + "left": "issue:45381", + "right": "issue:46693", + "accept": false, + "reason": "Qwen2.5-VL video position IDs vs FA2 long-context performance regression. Different symptoms and code paths." + }, + { + "left": "issue:41720", + "right": "issue:44991", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping vs tokenizer loading regression for est-roberta. Unrelated." + }, + { + "left": "issue:39921", + "right": "issue:44936", + "accept": false, + "reason": "Gemma3N special-token projection error vs trainer.evaluate after train. No meaningful overlap." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Sliding-window bug in Qwen3-MoE vs trust_remote_code AutoConfig missing fields. Different functionality." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio fbank feature extraction issue vs add-new-model-like failure in TOKENIZER_MAPPING_NAMES. Unrelated." + }, + { + "left": "issue:43611", + "right": "issue:46665", + "accept": false, + "reason": "Both concern model loading, but one is a base_model_prefix regression and the other is inconsistent from_pretrained behavior inside Python classes. Not the same concrete bug." + }, + { + "left": "issue:46829", + "right": "issue:46832", + "accept": false, + "reason": "VideoPrism output field mismatch vs RT-DETR forward TypeError from feature-level count. Different model heads and failures." + }, + { + "left": "issue:44008", + "right": "issue:46858", + "accept": false, + "reason": "Gemma3n audio_mel_mask name collision vs static-cache seq-length tensor arithmetic on non-CUDA backends. Unrelated." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset vs GenerationMixin import breakage. Different APIs and root causes." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "LayoutLMv2 tokenizer crash on NER/padding vs Perceiver non-default resolution failure. Only broad vision/tokenizer similarity." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "BigBird special-token handling vs DAC from_latents/STE mismatch. Different model internals and fixes." + }, + { + "left": "issue:44863", + "right": "issue:47458", + "accept": false, + "reason": "NemotronH checkpoint loading failure vs GPT2 c_proj weight initialization bug. Distinct model families and failure modes." + }, + { + "left": "issue:46650", + "right": "issue:47405", + "accept": false, + "reason": "Non-idempotent save_pretrained rename bug vs silent model_type mismatch conversion skip. Both serialization-related, but not the same defect." + }, + { + "left": "issue:44991", + "right": "issue:47020", + "accept": false, + "reason": "Both are tokenizer loading issues, but for different models and with different failure points. Not a duplicate pair." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Both hit _patch_mistral_regex, but one is an offline-mode network call bug and the other is an AttributeError on tokenizer shape. Same helper, different concrete bugs." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Both mention trust_remote_code / v5 loading regressions, but one is removal of is_torch_fx_available and the other is broader remote-code breakage. Too broad to merge." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "SigLIP2 doc/example errors vs Kosmos2.5 OCR index error. Different scopes and fixes." + }, + { + "left": "issue:45987", + "right": "issue:46829", + "accept": false, + "reason": "StaticCache.get_seq_length return-type bug vs VideoPrism hidden_states output mismatch. Not the same code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 21, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:43322", + "issue:43335", + "issue:43421", + "issue:43550", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43618", + "issue:43643", + "issue:43644", + "issue:43697", + "issue:43716", + "issue:43749", + "issue:43756", + "issue:43782", + "issue:43818", + "issue:44008", + "issue:44186", + "issue:44442", + "issue:44448", + "issue:44479", + "issue:44534", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44811", + "issue:44877", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45183", + "issue:45479", + "issue:45563", + "issue:45584", + "issue:45685", + "issue:45834", + "issue:45987", + "issue:46291", + "issue:46294", + "issue:46463", + "issue:46512", + "issue:46752", + "issue:46932", + "issue:47276", + "issue:47337", + "issue:47458" + ], + "soft_pairs": [ + "issue:43643|issue:44877", + "issue:43335|issue:43756", + "issue:44448|issue:44671", + "issue:45987|issue:46932", + "issue:44186|issue:44442", + "issue:43716|issue:44008", + "issue:43322|issue:45685", + "issue:44991|issue:46291", + "issue:46294|issue:46463", + "issue:44805|issue:44936", + "issue:43618|issue:43697", + "issue:43606|issue:45072", + "issue:45584|issue:45834", + "issue:43550|issue:45072", + "issue:44479|issue:45834", + "issue:43782|issue:44977", + "issue:45183|issue:45479", + "issue:46752|issue:47337", + "issue:43644|issue:43749", + "issue:45563|issue:47276", + "issue:43582|issue:46512", + "issue:43577|issue:43818", + "issue:43421|issue:44811", + "issue:44610|issue:45072", + "issue:44534|issue:47458" + ], + "prepared_review_unit_hash": "7cbac71cea774470ef8835aacdcff31971adbd1fb5a9ef69752915106b64a0b6", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 25, + "serialized_chars": 34764, + "estimated_input_tokens": 8691, + "estimated_eval_tokens": 17638 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the issues cover unrelated models, tokenizers, config loading, dtype/device mismatches, and warning/message regressions. None of the soft-similarity pairs look like the same underlying bug, so all are rejected as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue is safe here; the items are not one duplicate family but many unrelated bugs across different subsystems and models.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue for deduping stands out, because the soft links are mostly false positives and no item can represent the whole set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both are config-loading failures, but one is missing fields with trust_remote_code and the other is strict granite_speech config loading; different bugs." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and Smollm3 RoPE-layer dropping are unrelated model-architecture bugs." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Pegasus v4/v5 output drift and CamemBERT masked-LM prediction regressions affect different models and code paths." + }, + { + "left": "issue:45987", + "right": "issue:46932", + "accept": false, + "reason": "StaticCache seq-length return type and Gemma4 quantized embedder dtype casting are different cache vs quantization issues." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "LayoutLMv2 tokenization crashes and AutoTokenizer loading FastSpeech2ConformerTokenizer fail in different tokenizer paths." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Mistral-3 image-preprocessor dtype mismatch and Gemma3n variable-name collision are unrelated failures." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next loading segfaults and MPS histogram_mps int errors are different subsystems and error modes." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading for est-roberta and kernels version/revision import errors are unrelated." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen3.5-9B video chat-template mismatch and T5 decoder-input validation are different model-specific bugs." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask/index shape errors and trainer.evaluate() failing after train are not the same underlying issue." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "Missing CLIPOutput attentions and RTDetrV2 output drift are distinct regressions." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "CPU offload device mismatch in bark-small and bfloat16 dtype mismatches in SwitchTransformers/TimmWrapperModel are different problems." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription failure after align_special_tokens and Kosmos2.5 long-OCR index errors do not share a concrete code path." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "torch.compile/SDPA breakage in Bamba-9B-v2 is unrelated to dtype mismatch inference bugs." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Qwen VL video-input regression and Kosmos2.5 OCR index errors are different model/input pipelines." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Both mention Qwen, but one is a weight_only loading error and the other is a flash-attention generation bug; not the same defect." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Input validation messaging across models and single-label-classification zero-loss behavior are different bug classes." + }, + { + "left": "issue:46752", + "right": "issue:47337", + "accept": false, + "reason": "apply_chat_template empty-list crash and bbox-coordinate flaky IndexError are unrelated." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Non-persistent buffer junk filling and FSDP CPU RAM efficient loading are separate v5 regressions with different code paths." + }, + { + "left": "issue:45563", + "right": "issue:47276", + "accept": false, + "reason": "A stale num_return_sequences warning and a misleading synced_gpus warning are different warning-message issues." + }, + { + "left": "issue:43582", + "right": "issue:46512", + "accept": false, + "reason": "Apple Silicon caching_allocator_warmup TypeError and from_config dtype/config staleness are unrelated." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "BLIP2 dtype propagation and Video-LLaVA temporal-attention bugs affect different models and implementations." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Special-token post-processor updates and Whisper batch_decode skip_special_tokens handling are different tokenizer/processor behaviors." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor image-size mismatch and bfloat16 dtype mismatch are not the same bug." + }, + { + "left": "issue:44534", + "right": "issue:47458", + "accept": false, + "reason": "Non-persistent buffer junk and GPT2 c_proj weight initialization are both initialization-related but not the same concrete issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 22, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:42175", + "issue:42915", + "issue:43262", + "issue:43388", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43645", + "issue:43653", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43825", + "issue:43931", + "issue:43994", + "issue:44291", + "issue:44589", + "issue:44671", + "issue:44857", + "issue:44912", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45072", + "issue:45081", + "issue:45325", + "issue:45381", + "issue:45701", + "issue:46032", + "issue:46285", + "issue:46665", + "issue:46697", + "issue:46728", + "issue:46846", + "issue:46851", + "issue:47217", + "issue:47254", + "issue:47276", + "issue:47332", + "issue:47333", + "issue:47337" + ], + "soft_pairs": [ + "issue:34634|issue:36331", + "issue:28218|issue:29159", + "issue:43540|issue:45381", + "issue:43720|issue:45072", + "issue:43645|issue:46665", + "issue:43749|issue:44291", + "issue:47276|issue:47337", + "issue:43742|issue:45020", + "issue:42175|issue:45081", + "issue:43931|issue:46846", + "issue:43825|issue:43994", + "issue:43388|issue:46032", + "issue:47254|issue:47333", + "issue:42915|issue:47333", + "issue:46697|issue:46728", + "issue:44671|issue:45701", + "issue:43782|issue:45325", + "issue:43577|issue:43582", + "issue:47254|issue:47332", + "issue:44589|issue:45042", + "issue:36010|issue:43645", + "issue:43262|issue:47217", + "issue:46851|issue:47337", + "issue:43262|issue:43819", + "issue:43582|issue:44857", + "issue:43653|issue:44912", + "issue:36010|issue:36331", + "issue:45003|issue:46285" + ], + "prepared_review_unit_hash": "f76093db5cd51352d606fd8d5c239a4483e319520842f93c60a23dad11adead7", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36276, + "estimated_input_tokens": 9069, + "estimated_eval_tokens": 18394 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous and does not look like a duplicate set overall. Most soft pairs are only superficially similar (same subsystem or version-regression framing) but point to different code paths, models, or error modes.", + "confidence": 0.61, + "canonical_issue_reason": "Issue 45701 is the broadest tokenizer-regression report in the set and the best fit for the small tokenizer-related subset, but the cluster is not a coherent duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "45701 is the most general issue here; it best represents the only clearly overlapping theme (tokenization/version behavior), though it is still not a true canonical for the full cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "Different features and failures: BarkProcessor voice preset handling vs CustomTrainer.compute_loss API mismatch." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": false, + "reason": "Both concern tokenizers, but the described bugs differ: extra space after added tokens vs slow/fast tokenizer inconsistency." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Both are video-related, but they affect different Qwen models and different failure modes (ValueError vs wrong position_ids)." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "Unrelated load/inference bugs: BitNet packed-weight unpacking vs dtype mismatches in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43645", + "right": "issue:46665", + "accept": false, + "reason": "Both mention from_pretrained/model initialization, but one is a Jupyter custom-model regression and the other is behavior inside Python classes; not the same bug." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Different loading paths and root causes: FSDP CPU RAM efficient loading vs init_empty_weights getting an unexpected _is_hf_initialized argument." + }, + { + "left": "issue:47276", + "right": "issue:47337", + "accept": false, + "reason": "Completely unrelated: a warning message issue vs a flaky bounding-box test failure." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "A model-specific key error is not the same as the broad remote_code breakage; different underlying failures." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "TensorFlow backend packaging vs Mistral regex patch crash; no shared code-path bug." + }, + { + "left": "issue:43931", + "right": "issue:46846", + "accept": false, + "reason": "Different Qwen3-VL loading/shape issues and different mechanics (weight mismatch vs tensor-parallel plan omissions)." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message wording vs incorrect SigLIP2 outputs; one is messaging, the other is model behavior." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Different training/metrics and generation-cache bugs; they do not share the same underlying code path." + }, + { + "left": "issue:47254", + "right": "issue:47333", + "accept": false, + "reason": "Both involve training/offloading, but one is checkpointing with PEFT+DeepSpeed and the other is disk-offload save failure." + }, + { + "left": "issue:42915", + "right": "issue:47333", + "accept": false, + "reason": "Different Qwen models and different problems: FP8 config failure vs offloaded model save failure." + }, + { + "left": "issue:46697", + "right": "issue:46728", + "accept": false, + "reason": "Both are image-processing bugs, but one is over-padding and the other is patch-count calculation; not the same defect." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "A CamemBERT masked-LM regression is not the same as a generic tokenizer-version change report." + }, + { + "left": "issue:43782", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen2.5-VL issues: weight_only loading error vs rope index scaling for still images." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Blip2 dtype propagation and Apple Silicon allocator warmup are unrelated." + }, + { + "left": "issue:47254", + "right": "issue:47332", + "accept": false, + "reason": "Different offloading/training bugs: gradient checkpointing errors vs inability to save disk-offloaded Qwen3-VL." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Different domains: missing float8 storage type vs PIL image processors unnecessarily requiring torchvision." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "Both are import/loading regressions, but one is GenerationMixin import failure and the other is notebook-defined custom model initialization." + }, + { + "left": "issue:43262", + "right": "issue:47217", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting and VL image-placeholder failures are different multimodal bugs." + }, + { + "left": "issue:46851", + "right": "issue:47337", + "accept": false, + "reason": "Variable-length position_id handling is unrelated to the bounding-box flaky test." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Different model/pipeline bugs: audio processor chat template defaults vs DAC.from_latents mismatch." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "Apple Silicon allocator TypeError is unrelated to LwDetrImageLoss AMP crash." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and MXFP4 quantization fallback are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "ImportError for GenerationMixin and CustomTrainer compute_loss signature mismatch are separate API breakages." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access in modeling_utils is not the same as the CLIPTextModel structural inconsistency." + } + ] + }, + "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": 28, + "review_unit_index": 23, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28370", + "issue:29159", + "issue:38175", + "issue:42175", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43550", + "issue:43638", + "issue:43696", + "issue:43697", + "issue:43742", + "issue:43746", + "issue:43819", + "issue:43825", + "issue:43827", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:44060", + "issue:44291", + "issue:44368", + "issue:44479", + "issue:44610", + "issue:44811", + "issue:44821", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45042", + "issue:45081", + "issue:45084", + "issue:45372", + "issue:45406", + "issue:45538", + "issue:45859", + "issue:46175", + "issue:46209", + "issue:46291", + "issue:46326", + "issue:46932", + "issue:47050" + ], + "soft_pairs": [ + "issue:43531|issue:45042", + "issue:46175|issue:46932", + "issue:45372|issue:46326", + "issue:43550|issue:44912", + "issue:42175|issue:46291", + "issue:43638|issue:44936", + "issue:42915|issue:44821", + "issue:43819|issue:43927", + "issue:43257|issue:43329", + "issue:43452|issue:44991", + "issue:45081|issue:45406", + "issue:43257|issue:43696", + "issue:43825|issue:43827", + "issue:43526|issue:43697", + "issue:44060|issue:44368", + "issue:43854|issue:44977", + "issue:45859|issue:47050", + "issue:45084|issue:45538", + "issue:28370|issue:29159", + "issue:43906|issue:46209", + "issue:42915|issue:43742", + "issue:38175|issue:44610", + "issue:43746|issue:44811", + "issue:43329|issue:44479", + "issue:44291|issue:45042", + "issue:43262|issue:44811", + "issue:42490|issue:43257" + ], + "prepared_review_unit_hash": "ce340248d0e80cbfb7e912699fab3f42bc167ef0c0e2ebf9a61ced9257f6726a", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 36203, + "estimated_input_tokens": 9051, + "estimated_eval_tokens": 18358 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly heterogeneous one-off issues. Only one soft duplicate looks plausible: the Qwen video-input regression pair. Everything else appears to be separate bugs or unrelated subsystems.", + "confidence": 0.78, + "canonical_issue_reason": "issue:44479 is the best umbrella report for the only clear duplicate pair: it describes the broader v5.3.0 video-input regression across multiple Qwen VL/MoE models, while issue:43329 reads like a narrower root-cause report in the multimodal token-count path.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the strongest standalone issue in the cluster because it is broad, regression-framed, and captures the user-visible impact across several models.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Different bugs: Qwen3-MoE sliding-window logic vs PIL image processors incorrectly requiring torchvision." + }, + { + "left": "issue:46175", + "right": "issue:46932", + "accept": false, + "reason": "Same model family, but different code paths: a huge one-hot position-embedding allocation vs quantized vision/audio dtype casting." + }, + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "Both involve mistral_common, but one is an import error during Gemma 4 processor loading and the other is continue_final_message tokenizer behavior." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "Unrelated issues: torch.compile/SDPA failure for Bamba vs MXFP4 quantization fallback for git-oss-20b." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Different dependency failures: missing TensorFlow in a torch install vs kernels v0.15 import-time version/revision requirements." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "Both are training-related, but one is a ZeRO-3/non-pretrained Bert IndexError and the other is evaluate() failing after train(); not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Completely different subsystems: Qwen3-MoE FP8 config failure vs AutoImageProcessor loading from URL." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "Different models and failures: DAC latent reconstruction mismatch vs DiaConfig losing custom token IDs after save/load." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "One is Qwen3-MoE weight conversion with accelerate+deepspeed; the other is a multimodal token-count video-branch bug." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both concern tokenizer/model loading, but one is gguf_file support for auto loaders and the other is a specific v5 tokenizer regression for EMBEDDIA/est-roberta." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Both are loading-time errors, but they affect different components: Mistral tokenizer regex patching vs Gemma4Processor tokenization state." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "No shared underlying bug: Qwen3-MoE conversion issue vs GPT-oss-20b CUDA OOM." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Related to pipeline removal, but one is a runtime error message and the other is stale docs; not the same concrete bug/change." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "Different image/model bugs: BeitImageProcessorFast label reduction vs RTDetrV2 output divergence in v5." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both mention tied-weights warnings, but the concrete symptoms and models differ; not clearly the same code path." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "Different model-specific inference/loading problems: GLM-4.7-Flash unit-test loading vs Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:45859", + "right": "issue:47050", + "accept": false, + "reason": "Unrelated bugs: CPU-offload incompatibility in Qwen3_5Moe rotary embedding vs CanineModel fp16 ONNX attention dtype mismatch." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "No overlap beyond being bugs: template compilation TypeError vs CLIPTokenizer model_max_length sentinel value." + }, + { + "left": "issue:28370", + "right": "issue:29159", + "accept": false, + "reason": "Both are tokenizer-related, but one is missing vocab_file for custom SentencePiece models and the other is slow/fast tokenizer inconsistency in general." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "Different reports: an isolated reproduction of an earlier issue vs an FP8 inference failure on B200 hardware." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "Different model-loading failures with no shared code path: Qwen3-MoE FP8 config vs MobileLLM-125M key error." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Both are vision-model complaints, but one is zero probabilities in SigLIP2 and the other is an image-size mismatch in OmDet-Turbo." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "Different functionality: PEFT adapter loading for GraniteSpeech vs Whisper batch_decode skip_special_tokens handling." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": true, + "reason": "These look like the same video-input regression: both are about broken multimodal/video handling in Qwen VL/MoE models, and 43329 provides a likely root cause in the video token-count path while 44479 is the broader regression report." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "Unrelated issues: init_empty_weights/_is_hf_initialized argument handling vs PIL image processors requiring torchvision." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Different audio/processor bugs: apply_chat_template sampling-rate default vs Whisper batch_decode special-token skipping." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "No shared bug: local-path trailing-slash import warning vs Qwen3-MoE weight conversion with accelerate+deepspeed." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative overall. The only accepted soft pair, issue:43329 vs issue:44479, is plausibly the same Qwen video-input regression at different levels of specificity. The other pairwise rejections are appropriately cautious and do not overclaim duplication." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 24, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28282", + "issue:29159", + "issue:29870", + "issue:33357", + "issue:34567", + "issue:34689", + "issue:39692", + "issue:40990", + "issue:43257", + "issue:43329", + "issue:43425", + "issue:43450", + "issue:43454", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43577", + "issue:43643", + "issue:43650", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43906", + "issue:43931", + "issue:44291", + "issue:44387", + "issue:44554", + "issue:44589", + "issue:44617", + "issue:44756", + "issue:44821", + "issue:44912", + "issue:44987", + "issue:45081", + "issue:45459", + "issue:45538", + "issue:45600", + "issue:45759", + "issue:45803", + "issue:46291", + "issue:46701", + "issue:47020", + "issue:47145", + "issue:47332", + "issue:47333", + "issue:47458" + ], + "soft_pairs": [ + "issue:43643|issue:45759", + "issue:43329|issue:45600", + "issue:39692|issue:44912", + "issue:44387|issue:44617", + "issue:34567|issue:40990", + "issue:43526|issue:44554", + "issue:43329|issue:43450", + "issue:44589|issue:47333", + "issue:29159|issue:34567", + "issue:45081|issue:47020", + "issue:43425|issue:46291", + "issue:43257|issue:43931", + "issue:44291|issue:44821", + "issue:43756|issue:47458", + "issue:44987|issue:45538", + "issue:43525|issue:46291", + "issue:44912|issue:47145", + "issue:43540|issue:46701", + "issue:43742|issue:44291", + "issue:45459|issue:45803", + "issue:43454|issue:43577", + "issue:43749|issue:44756", + "issue:43531|issue:44821", + "issue:28282|issue:34689", + "issue:43650|issue:43906", + "issue:33357|issue:34567", + "issue:28282|issue:29870", + "issue:44589|issue:47332" + ], + "prepared_review_unit_hash": "4fd66c353fb7957d466f257224a4c112f2441b9a5a38f714cfa3568f74ac5b25", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35606, + "estimated_input_tokens": 8902, + "estimated_eval_tokens": 18060 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a false-positive mix of unrelated Transformers issues spanning config loading, tokenizers, video/multimodal processing, quantization, training state, and model-specific regressions. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.89, + "canonical_issue_reason": "If forced to pick a single anchor, issue 43643 is the closest because it is a broad config-loading regression, but it does not actually subsume the rest of the cluster, so there is no real canonical duplicate here.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43643 is the most representative standalone report in this set, but only as a loose loader/config example; it is not a good global duplicate center for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "Related config-loading surface, but different bugs: missing returned fields from AutoConfig vs not unwrapping text_config in AutoModelForCausalLM." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "One is a video token counting undefined-variable bug; the other is stale auto_mappings references to removed configs." + }, + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Documentation example errors are unrelated to an MXFP4 quantization fallback bug." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both are memory-related, but one is an int4 reserved-memory regression and the other is a model-specific CUDA OOM." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "Trainer token-count bookkeeping is unrelated to high perplexity on a benchmark." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "Image-processor label reduction bug and MPS attention correctness issue are different code paths and failures." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Both mention video processing, but one is an undefined helper in multimodal token counting and the other is a batched-shape bug in processors." + }, + { + "left": "issue:44589", + "right": "issue:47333", + "accept": false, + "reason": "Storage lookup failure and disk-offload save failure are distinct loading/saving problems." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency is unrelated to TrainerState token accounting." + }, + { + "left": "issue:45081", + "right": "issue:47020", + "accept": false, + "reason": "Different tokenizer internals: backend_tokenizer attribute access vs passing the wrong type to Unigram." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch version incompatibility and kernels package import requirements are separate dependency issues." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "MoE weight conversion under accelerate/deepspeed is not the same as a Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "init_empty_weights argument regression is unrelated to loading AutoImageProcessor from a URL." + }, + { + "left": "issue:43756", + "right": "issue:47458", + "accept": false, + "reason": "RoPE layer-count/config mismatch is not the same as GPT2 c_proj initialization being wrong." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Version-specific model loading failure and CLIPTokenizer max_length misuse are different bugs." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Missing pad_token_id on Llama4Config is unrelated to pipelines import failing on kernels metadata." + }, + { + "left": "issue:44912", + "right": "issue:47145", + "accept": false, + "reason": "MXFP4 fallback-to-bf16 loading bug and FP8 frozen-checkpoint weight_scale omission are different quantization issues." + }, + { + "left": "issue:43540", + "right": "issue:46701", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError is unrelated to DiffusionGemmaGenerationOutput pipeline behavior." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "MobileLLM key error is model-specific; init_empty_weights is a broader API regression." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Both involve exception handling, but they are different functions and different failure modes." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "Weight tying causing garbage generation is unrelated to BLIP2 dtype propagation staying float32." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "FSDP CPU RAM-efficient loading and disabling mmap on Strix Halo are separate memory-management issues." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor URL loading are unrelated." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "Missing PyTorch installation error is not the same as a model-loading regression for Llama 3.2 Vision." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "A vague placeholder issue is not a duplicate of a concrete reproduction report." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS bus error in CLIP loading is unrelated to TrainerState token counting." + }, + { + "left": "issue:28282", + "right": "issue:29870", + "accept": false, + "reason": "ImportError for missing torch and BART generate decoder_attention_mask bug are different failures." + }, + { + "left": "issue:44589", + "right": "issue:47332", + "accept": false, + "reason": "Float8 storage lookup failure and disk-offloaded model save failure are separate code paths." + }, + { + "left": "issue:43749", + "right": "issue:47333", + "accept": false, + "reason": "FSDP CPU RAM efficient loading and saving a disk-offloaded Qwen3 model are different offloading 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": 28, + "review_unit_index": 25, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:24643", + "issue:27957", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:33666", + "issue:34567", + "issue:35308", + "issue:36331", + "issue:38175", + "issue:39379", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43404", + "issue:43452", + "issue:43540", + "issue:43582", + "issue:43697", + "issue:43749", + "issue:43824", + "issue:44062", + "issue:44479", + "issue:44488", + "issue:44554", + "issue:44560", + "issue:44568", + "issue:44857", + "issue:44898", + "issue:44936", + "issue:44991", + "issue:45081", + "issue:45127", + "issue:45137", + "issue:45356", + "issue:45357", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45538", + "issue:46421", + "issue:46424", + "issue:46693", + "issue:46821", + "issue:46832", + "issue:46869", + "issue:47254", + "issue:47332" + ], + "soft_pairs": [ + "issue:44479|issue:46693", + "issue:42175|issue:43824", + "issue:44568|issue:45356", + "issue:42915|issue:45538", + "issue:46421|issue:46424", + "issue:43582|issue:47254", + "issue:24643|issue:29159", + "issue:44568|issue:45488", + "issue:44936|issue:46821", + "issue:43697|issue:45357", + "issue:43404|issue:45127", + "issue:45538|issue:46869", + "issue:28282|issue:29127", + "issue:45137|issue:45405", + "issue:41720|issue:43540", + "issue:44062|issue:44991", + "issue:44857|issue:46832", + "issue:43540|issue:44479", + "issue:43540|issue:44560", + "issue:39379|issue:44554", + "issue:42915|issue:43749", + "issue:42915|issue:47332", + "issue:33666|issue:35308", + "issue:44479|issue:45381", + "issue:43452|issue:43582", + "issue:34567|issue:36331", + "issue:27957|issue:28282", + "issue:44062|issue:45081", + "issue:38175|issue:44898", + "issue:42175|issue:44488" + ], + "prepared_review_unit_hash": "b96053d58aab1dbc773c08a24bcb9a4d4d28e7d5c4ed10f436fe25559116a195", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36761, + "estimated_input_tokens": 9191, + "estimated_eval_tokens": 18638 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue cluster. I only found one credible duplicate edge: the Qwen2.5-VL video-input regression report aligns with the broader Qwen video-input regression issue. The rest look like separate bugs across different subsystems/models.", + "confidence": 0.89, + "canonical_issue_reason": "issue:44479 is the best anchor because it states the broad video-input regression affecting multiple Qwen multimodal variants, and it cleanly subsumes the more specific qwen2.5-vl video-position report.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the strongest representative issue in the cluster: it has the clearest regression framing, the broadest model scope, and matches the one accepted duplicate edge.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44479", + "right": "issue:46693", + "accept": false, + "reason": "Both mention video/long-context performance, but one is a Qwen multimodal video-input regression and the other is an FA2 long-context slowdown. Different bugs and code paths." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Backend packaging vs missing-model import error; unrelated failure modes." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different tokenizers and different code paths. Too broad to be the same bug." + }, + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Qwen3Moe FP8 failure and CLIP tokenizer max-length handling are unrelated." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both are cache-related, but one is lazy init causing recompilation and the other is max_cache_length being ignored outside quantized cache. Separate issues." + }, + { + "left": "issue:43582", + "right": "issue:47254", + "accept": false, + "reason": "Different distributed-training failures: Apple Silicon warmup TypeError vs PEFT/DeepSpeed checkpointing error." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed training runtime error and tokenizer consistency are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Both involve tokenizers, but one is special-token duplication and the other is LlamaTokenizer pre-tokenizer overriding. Different underlying bugs." + }, + { + "left": "issue:44936", + "right": "issue:46821", + "accept": false, + "reason": "Trainer lifecycle bug vs TimesFM window_size AttributeError; not the same problem." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "RT-DETR output mismatch and Qwen3.5 save_pretrained key corruption are different model behaviors and code paths." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Both mention tied weights, but one is a specific lm_head tying bug and the other is LoRA merge collapse with extended vocab. Not the same concrete defect." + }, + { + "left": "issue:45538", + "right": "issue:46869", + "accept": false, + "reason": "Tokenizer max_length issue vs YOLO multi-GPU device-map failure; unrelated." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Missing PyTorch import error is unrelated to LayoutLMv3 box-validation messaging." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO-3 deque error and PEFT version bump are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both are Qwen model failures, but one is auto device mapping CUDA assert and the other is video-input processing in Qwen3OmniMoe. Different bugs." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Generic AddedToken constructor conflict vs a specific tokenizer-loading failure for EMBEDDIA/est-roberta." + }, + { + "left": "issue:44857", + "right": "issue:46832", + "accept": false, + "reason": "Different object-detection bugs: AMP/CUDA loss crash vs RT-DETR forward shape mismatch." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": false, + "reason": "Both concern Qwen video inputs, but 43540 is a specific Qwen3OmniMoe ValueError while 44479 is a broader multi-model regression. Not enough evidence they are the same concrete bug." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-input failures, but they involve different Qwen model families and different errors." + }, + { + "left": "issue:39379", + "right": "issue:44554", + "accept": false, + "reason": "Mllama mask handling and MPS attention correctness are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:43749", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and FSDP CPU RAM efficient loading are different distributed/runtime issues." + }, + { + "left": "issue:42915", + "right": "issue:47332", + "accept": false, + "reason": "FP8 config failure and disk-offload save failure are unrelated." + }, + { + "left": "issue:33666", + "right": "issue:35308", + "accept": false, + "reason": "Both are multi-GPU training issues, but one is Qwen2-VL-specific and the other is iterable-dataset length mismatch. Different root causes." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Strong match: both describe the same Qwen2.5-VL video-input regression in transformers 5.3.0, with 45381 giving the specific vision_position_ids symptom." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "gguf_file tokenizer/model loading breakage is unrelated to the Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token counting and custom compute_loss signature mismatch are different trainer API issues." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention support for XLM-Roberta is unrelated to missing PyTorch import detection." + }, + { + "left": "issue:44062", + "right": "issue:45081", + "accept": false, + "reason": "AddedToken keyword collision and Mistral regex patch crashing on backend_tokenizer are different tokenizer bugs." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Siglip2 zero probabilities and Perceiver interpolation failure are unrelated vision-model issues." + }, + { + "left": "issue:42175", + "right": "issue:44488", + "accept": false, + "reason": "TensorFlow packaging/backend omission and a specific tokenizer load failure are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative. The lone accepted edge (44479 \u2194 45381) is clearly supported by the issue titles, and the rejected soft edges are framed as distinct bugs without overclaiming. No mergeability or PR-pair concerns here." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 26, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:28370", + "issue:29870", + "issue:43257", + "issue:43452", + "issue:43493", + "issue:43540", + "issue:43577", + "issue:43716", + "issue:43742", + "issue:43756", + "issue:43906", + "issue:43994", + "issue:44466", + "issue:44488", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44737", + "issue:44792", + "issue:44821", + "issue:44849", + "issue:44877", + "issue:44964", + "issue:44987", + "issue:45005", + "issue:45072", + "issue:45325", + "issue:45405", + "issue:45600", + "issue:45672", + "issue:45674", + "issue:45701", + "issue:45823", + "issue:45907", + "issue:46399", + "issue:46463", + "issue:46491", + "issue:46512", + "issue:46693", + "issue:46701", + "issue:46821", + "issue:46899", + "issue:47276", + "issue:47332", + "issue:47333", + "issue:47405" + ], + "soft_pairs": [ + "issue:43742|issue:44589", + "issue:43493|issue:43994", + "issue:43577|issue:46512", + "issue:28282|issue:28370", + "issue:45672|issue:45674", + "issue:43742|issue:47333", + "issue:44792|issue:44964", + "issue:43716|issue:46701", + "issue:44849|issue:45701", + "issue:44987|issue:45907", + "issue:43756|issue:45325", + "issue:45823|issue:46463", + "issue:43756|issue:46399", + "issue:44737|issue:46463", + "issue:46463|issue:47276", + "issue:43257|issue:43756", + "issue:45405|issue:45600", + "issue:43906|issue:46693", + "issue:44625|issue:44877", + "issue:43540|issue:46693", + "issue:44466|issue:45005", + "issue:43742|issue:47332", + "issue:43577|issue:45072", + "issue:43257|issue:47405", + "issue:44568|issue:46491", + "issue:43452|issue:44488", + "issue:24643|issue:29870", + "issue:46463|issue:46899", + "issue:44821|issue:46821" + ], + "prepared_review_unit_hash": "451a11cea3f8d5838425b716d2952ae0d5c884f2b11e55056297ee597f6b4659", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36624, + "estimated_input_tokens": 9156, + "estimated_eval_tokens": 18568 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous model-loading / dtype / tokenizer bug reports, with only two plausible duplicate mini-pairs: SigLIP2 output-quality issues and tied-weights handling in v5. Most other soft matches are superficial subsystem overlap rather than the same bug.", + "confidence": 0.66, + "canonical_issue_reason": "issue:44466 is the clearest, most root-cause-oriented report in the only broader duplicate family here: v5 tied-weight serialization varying by device. It is the best anchor for the tied-weights subgroup.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44466 is the strongest overall representative of the cluster\u2019s limited duplicate signal: it names the concrete tied-weights bug directly and is broad enough to subsume the closely related v5 tied-weights report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "Different failures: model loading key error vs Float8 storage lookup/serialization error. Same broad area, not the same bug." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Both point to SigLIP2 producing wrong outputs versus expected behavior/reference implementation; likely the same underlying model implementation bug." + }, + { + "left": "issue:43577", + "right": "issue:46512", + "accept": false, + "reason": "Related dtype theme, but one is BLIP2/Qformer staying float32 while the other is from_config leaving config.dtype stale. Different code paths." + }, + { + "left": "issue:28282", + "right": "issue:28370", + "accept": false, + "reason": "ImportError for missing PyTorch is unrelated to missing vocab_file for custom SentencePiece models." + }, + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Both involve 4-bit quantization, but one is an unsafe dtype cast in Gemma4 audio modules and the other is a skip-module/default-exclusion assertion bug." + }, + { + "left": "issue:43742", + "right": "issue:47333", + "accept": false, + "reason": "Different model families and different operations: loading MobileLLM vs saving a disk-offloaded Qwen3 model." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Both are multimodal-model issues, but they target different models and different failure surfaces (test generation vs loading)." + }, + { + "left": "issue:43716", + "right": "issue:46701", + "accept": false, + "reason": "Different models and bugs: Mistral-3 preprocessor dtype mismatch vs DiffusionGemma pipeline failure." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-state bug is unrelated to tokenization changes across versions." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Model loading regression vs repo-files revision forwarding API bug; not the same underlying change." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Both mention positional/RoPE logic, but they are different model-specific bugs with different symptoms and fixes." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 device placement mismatch and T5 decoder input validation messaging are unrelated." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "Different model subsystems: Smollm3 RoPE-layer count bug vs Qwen3.5 kernel registration failure." + }, + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "XLNet CPU allocation bug is unrelated to T5 decoder validation error handling." + }, + { + "left": "issue:46463", + "right": "issue:47276", + "accept": false, + "reason": "Both are message/warning quality issues, but they are different code paths and not the same bug." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MOE weight conversion under accelerate+deepspeed is unrelated to Smollm3 RoPE layer dropping." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "One is a version-pin/dependency issue; the other is stale auto_mappings entries breaking CI." + }, + { + "left": "issue:43906", + "right": "issue:46693", + "accept": false, + "reason": "Insufficient evidence of the same bug; one is a reproduction of another issue, the other is a performance regression in FA2 long-context generation." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "Both are config-related, but one is Qwen3.5 num_labels propagation and the other is strict config loading for granite_speech." + }, + { + "left": "issue:43540", + "right": "issue:46693", + "accept": false, + "reason": "Video-input handling in Qwen3OmniMoe is unrelated to the FA2 performance regression." + }, + { + "left": "issue:44466", + "right": "issue:45005", + "accept": true, + "reason": "Both describe v5 tied-weights handling problems; the shared underlying issue is serialization/weight-tie behavior changing in v5 for translation-style models." + }, + { + "left": "issue:43742", + "right": "issue:47332", + "accept": false, + "reason": "Different model-loading vs disk-offload save failure; no same code-path evidence." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both are dtype bugs, but they affect different model stacks and likely different fixes." + }, + { + "left": "issue:43257", + "right": "issue:47405", + "accept": false, + "reason": "Deepspeed conversion failure is not the same as silent conversion skipping when loading through the wrong Auto class." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Tokenizer regressions, but different behaviors and different models; not the same bug." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "Both are loading-related, but one is a gguf_file regression across AutoTokenizer/AutoModelForCausalLM and the other is a specific model load failure without shared code-path evidence." + }, + { + "left": "issue:24643", + "right": "issue:29870", + "accept": false, + "reason": "DeepSpeed training weight-shape error is unrelated to Bart generate() using decoder_attention_mask before assignment." + }, + { + "left": "issue:46463", + "right": "issue:46899", + "accept": false, + "reason": "T5 validation messaging and Gemma4 4-bit multimodal dtype casting are unrelated bugs." + }, + { + "left": "issue:44821", + "right": "issue:46821", + "accept": false, + "reason": "AutoImageProcessor URL-loading failure is unrelated to TimesFM window_size AttributeError." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The clustering summary is grounded in the packet and stays appropriately conservative. The two accepted soft-edge pairs are plausible from the issue titles alone, and the many rejections are well-justified by differing models, failure surfaces, or code paths. No overstatement stands out." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 27, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:33666", + "issue:36010", + "issue:38175", + "issue:40444", + "issue:42796", + "issue:43295", + "issue:43452", + "issue:43572", + "issue:43643", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43697", + "issue:43720", + "issue:43749", + "issue:43761", + "issue:43906", + "issue:43927", + "issue:44361", + "issue:44442", + "issue:44493", + "issue:44737", + "issue:44756", + "issue:44811", + "issue:44843", + "issue:44857", + "issue:44898", + "issue:44913", + "issue:44991", + "issue:45216", + "issue:45357", + "issue:45372", + "issue:45542", + "issue:45588", + "issue:45663", + "issue:45701", + "issue:45704", + "issue:45920", + "issue:46050", + "issue:46121", + "issue:46170", + "issue:46178", + "issue:46829", + "issue:46846", + "issue:47000", + "issue:47013", + "issue:47332", + "issue:47404" + ], + "soft_pairs": [ + "issue:42796|issue:43906", + "issue:43452|issue:44843", + "issue:45663|issue:46846", + "issue:44756|issue:47404", + "issue:43720|issue:44898", + "issue:43697|issue:46829", + "issue:44737|issue:45588", + "issue:44913|issue:46121", + "issue:43653|issue:44361", + "issue:44493|issue:46050", + "issue:38175|issue:43295", + "issue:45357|issue:45920", + "issue:44756|issue:47332", + "issue:44361|issue:44442", + "issue:46846|issue:47404", + "issue:40444|issue:44811", + "issue:43697|issue:45216", + "issue:44843|issue:44991", + "issue:43688|issue:43749", + "issue:45372|issue:46170", + "issue:44857|issue:47000", + "issue:43761|issue:47013", + "issue:43572|issue:43643", + "issue:43646|issue:43749", + "issue:45542|issue:46178", + "issue:33666|issue:36010", + "issue:43688|issue:45701", + "issue:43688|issue:45704", + "issue:43653|issue:43927" + ], + "prepared_review_unit_hash": "60691c649c04a7ea3122cf17d9e18271f63a9095be73541061bf08a0c45e4ff5", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36780, + "estimated_input_tokens": 9195, + "estimated_eval_tokens": 18646 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the soft pairs mostly link unrelated bug reports that share only broad themes like tokenizers, loading, or regressions. I would not merge any of the candidate pairs.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45920 is the strongest anchor: it\u2019s a broad, high-impact v5 tokenizer regression affecting multiple models, with a clear symptom and wide applicability.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best representative issue in this set because it is well-scoped, broadly relevant, and clearly describes a user-facing regression with multiple affected models.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42796", + "right": "issue:43906", + "accept": false, + "reason": "Different tokenizer bugs/models; one is BPE marker handling, the other is an isolated reproduction of a different prior issue." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both involve loading, but one is gguf_file handling and the other is an offline model_info call in _patch_mistral_regex." + }, + { + "left": "issue:45663", + "right": "issue:46846", + "accept": false, + "reason": "Different models and failure modes: FSDP2 KeyError in Gemma-4 vs Qwen3.5 tensor-parallel OOM/shape issues." + }, + { + "left": "issue:44756", + "right": "issue:47404", + "accept": false, + "reason": "Both are OOM-related, but mmap on Strix Halo and destination-GPU contiguous() transients are different code paths." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "Unrelated models and bugs; BitNet packed-weight loading vs Perceiver interpolation failure." + }, + { + "left": "issue:43697", + "right": "issue:46829", + "accept": false, + "reason": "Both report output mismatches, but the models and output fields differ; not the same underlying defect." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet CPU placement bug is unrelated to the flash_attention.py s_aux=None crash." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Both touch rope/config logic, but one is rotary_pct persistence and the other is list-type validation in a helper." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Different tokenizer implementations and error modes: special-token decode issue vs MLuke AttributeError on tasks." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Position-id warnings and quantized save/load breakage are not the same bug or code path." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "siglip2 zero probabilities and processor.tokenizer regression are separate vision/text issues." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Qwen3.5 visual-encoder save keys and broad AutoTokenizer wrong-ID regression are different problems." + }, + { + "left": "issue:44756", + "right": "issue:47332", + "accept": false, + "reason": "Both are offloading-related, but mmap OOM avoidance and saving a disk-offloaded model are distinct failures." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Different tokenizer models and failures: MLuke task AttributeError vs FastSpeech2Conformer loading error." + }, + { + "left": "issue:46846", + "right": "issue:47404", + "accept": false, + "reason": "Different subsystems: Qwen3.5 tensor-parallel plan vs device_map conversion-op OOM." + }, + { + "left": "issue:40444", + "right": "issue:44811", + "accept": false, + "reason": "IterableDataset multi-image finetuning failure is unrelated to Whisper batch_decode skip_special_tokens behavior." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained checkpoint corruption are unrelated regressions." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Offline model_info access and est-roberta tokenizer loading regression do not share the same concrete code path." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are separate bugs." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Gemma processor import failure and Windows ROCm torch.distributed dependency issue are different failures." + }, + { + "left": "issue:44857", + "right": "issue:47000", + "accept": false, + "reason": "Different loss functions/models and different crash causes: AMP float16 failure vs infeasible cost matrix." + }, + { + "left": "issue:43761", + "right": "issue:47013", + "accept": false, + "reason": "CLIP hidden_states regression and xLSTM return_last_states unpacking bug are unrelated." + }, + { + "left": "issue:43572", + "right": "issue:43643", + "accept": false, + "reason": "StableLmConfig pad_token_idx loss and trust_remote_code missing fields are different config bugs." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "General custom model initialization regression vs FSDP CPU RAM loading; separate issues." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both are backend-related, but one is undefined tf backend from tensorboard-only install and the other is unsupported tensorflow_text backend." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Qwen2-VL multi-GPU training and GenerationMixin import error are unrelated." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "Aux-loss normalization and tokenizer behavior changes across versions are separate bugs." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Aux-loss normalization has no concrete overlap with T5 apex.FusedRMSNorm memory leak." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "Both are tokenizer/config related, but BigBird special-token registration and DiaConfig custom token IDs 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": 28, + "review_unit_index": 28, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28370", + "issue:32090", + "issue:33666", + "issue:34567", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:42898", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43577", + "issue:43606", + "issue:43643", + "issue:43653", + "issue:43688", + "issue:43720", + "issue:43761", + "issue:43782", + "issue:43927", + "issue:44361", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44561", + "issue:44568", + "issue:44610", + "issue:44625", + "issue:44743", + "issue:44977", + "issue:45356", + "issue:45405", + "issue:45440", + "issue:45759", + "issue:46050", + "issue:46491", + "issue:46693", + "issue:46701", + "issue:46829", + "issue:46832", + "issue:46932", + "issue:47254", + "issue:47276" + ], + "soft_pairs": [ + "issue:43525|issue:46701", + "issue:43295|issue:45356", + "issue:44479|issue:44625", + "issue:44464|issue:44743", + "issue:46832|issue:46932", + "issue:43550|issue:44361", + "issue:43761|issue:46829", + "issue:43653|issue:44568", + "issue:41720|issue:44464", + "issue:43452|issue:43577", + "issue:33666|issue:36331", + "issue:44977|issue:46693", + "issue:28370|issue:34567", + "issue:44464|issue:44977", + "issue:43504|issue:43550", + "issue:44466|issue:46050", + "issue:43782|issue:46693", + "issue:42898|issue:46491", + "issue:32090|issue:33666", + "issue:45405|issue:47254", + "issue:43606|issue:43927", + "issue:43720|issue:43927", + "issue:43504|issue:43720", + "issue:43643|issue:44561", + "issue:43688|issue:45440", + "issue:43278|issue:43688", + "issue:43504|issue:44610", + "issue:39692|issue:43493", + "issue:44625|issue:47276", + "issue:43720|issue:44361", + "issue:43334|issue:45759" + ], + "prepared_review_unit_hash": "08164063efca208e52c9b9bd4ec6b20139609decf597ccdc4b8c3048acbedb21", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37338, + "estimated_input_tokens": 9335, + "estimated_eval_tokens": 18926 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are a heterogeneous mix of unrelated Transformers bugs/regressions (tokenizers, model loading, training, generation, vision, MoE, quantization). None of the soft pairs look like the same underlying bug or change, so no duplicates are accepted.", + "confidence": 0.91, + "canonical_issue_reason": "Issue #46693 is the most central by discussion volume and inbound references, with a broad user-facing regression impact; it is the best single representative despite the overall cluster being heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "#46693 is the strongest representative of the set because it is the most heavily discussed and clearly documented regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43525", + "right": "issue:46701", + "accept": false, + "reason": "Different model families and failure modes: missing `pad_token_id` on Qwen3-VL config vs a DiffusionGemma image-text pipeline bug." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Both are regressions, but one is about processor/tokenizer access in custom model code and the other is about Kimi-K2.5 codec/tokenization handling; different code paths." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Different bugs: video input regression across Qwen VL variants vs `num_labels` propagation into Qwen3.5 text config." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Both touch generation/caching, but one is chunked generation with compiled forward and the other is recurrent state reset in a modular Qwen3.5 file; not the same concrete bug." + }, + { + "left": "issue:46832", + "right": "issue:46932", + "accept": false, + "reason": "Unrelated areas: RT-DETR feature-level mismatch vs Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Completely different subsystems: torch.compile + SDPA in Bamba vs tokenizer AttributeError in MLukeTokenizer." + }, + { + "left": "issue:43761", + "right": "issue:46829", + "accept": false, + "reason": "Both involve hidden_states, but one is CLIPVisionModel omitting them and the other is VideoPrism classification output wiring; different implementation bugs." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer special-token issues, but they affect different tokenizers and different behaviors (mask token registration vs BOS/EOS insertion)." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Different failure modes: CUDA assert under auto device mapping for Qwen3 vs inconsistent outputs from compiled chunked generation." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "Different loading/serialization issues: gguf_file breaks AutoModel/AutoTokenizer loading vs Blip2 dtype not propagating to model/qformer." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Different trainer-related problems: Qwen2-VL multi-GPU training vs a custom `compute_loss` signature regression." + }, + { + "left": "issue:44977", + "right": "issue:46693", + "accept": false, + "reason": "Both mention generation and attention, but one is correctness for Qwen3.5 with flash-attention and the other is a performance regression with FA2 long-context generation." + }, + { + "left": "issue:28370", + "right": "issue:34567", + "accept": false, + "reason": "Unrelated: SentencePiece `vocab_file` attribute handling vs TrainerState token accounting not updating." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Different generation bugs with different triggers: compiled forward inconsistency vs flash-attention misgeneration." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Different model-loading vs compile/attention problems; no shared code-path bug." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both concern save/load behavior, but one is tied-weight serialization varying by device and the other is sinq quantization breaking after reload." + }, + { + "left": "issue:43782", + "right": "issue:46693", + "accept": false, + "reason": "Different root causes: Qwen3VL `from_pretrained(weight_only=True)` error vs a flash-attention long-context performance regression." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "Both are tokenization regressions, but one is a cleanup-space behavior change and the other is a CodeLlama leading-space round-trip bug." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Both involve training, but one is a `_gpu_broadcast_one` NoneType error and the other is Qwen2-VL multi-GPU training behavior." + }, + { + "left": "issue:45405", + "right": "issue:47254", + "accept": false, + "reason": "No overlap: an unreleased PEFT version bump vs a checkpointing error with PEFT + DeepSpeed ZeRO-3." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "Different bugs: CPU offload device mismatch in bark-small vs DiaConfig losing custom token IDs after save/load." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Different areas and symptoms: BitNet packed-weight handling during accelerate loading vs MLukeTokenizer task-time AttributeError." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Related only by `trust_remote_code`, but one is missing fields from `AutoConfig.from_pretrained` and the other is a v5 removal breaking remote-code models." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Both are MoE-related, but one is auxiliary-loss normalization and the other is a DeepSeekV3MoE implementation divergence." + }, + { + "left": "issue:43278", + "right": "issue:43688", + "accept": false, + "reason": "Different bugs: dtype mismatch between train/eval for embeddings vs auxiliary-loss normalization in MoE models." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "Different vision-model bugs: legacy-field loading failure in BEiT vs OmDet-Turbo processor/model image-size mismatch." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Both mention SigLIP2, but one is a docs/example issue with model/processor mismatch and quantization failure, while the other is an implementation discrepancy vs JAX." + }, + { + "left": "issue:44625", + "right": "issue:47276", + "accept": false, + "reason": "Completely unrelated: Qwen3.5 `num_labels` propagation vs a misleading `synced_gpus` warning in continuous batching." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "Same as above: unrelated BitNet loading issue vs MLukeTokenizer task error; not mergeable as one bug fix." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Both are Qwen multimodal config issues, but one is a missing `pad_token_id` attribute and the other is `from_config` failing to unwrap `text_config`; different fixes and code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 29, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28218", + "issue:33357", + "issue:39692", + "issue:42796", + "issue:43257", + "issue:43295", + "issue:43334", + "issue:43493", + "issue:43525", + "issue:43531", + "issue:43572", + "issue:43577", + "issue:43611", + "issue:43646", + "issue:43825", + "issue:43927", + "issue:43994", + "issue:44291", + "issue:44361", + "issue:44466", + "issue:44484", + "issue:44488", + "issue:44493", + "issue:44514", + "issue:44554", + "issue:44568", + "issue:44704", + "issue:44734", + "issue:44843", + "issue:44964", + "issue:44977", + "issue:45290", + "issue:45405", + "issue:45491", + "issue:45812", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46192", + "issue:46209", + "issue:46365", + "issue:46489", + "issue:46612", + "issue:46846", + "issue:47332", + "issue:47404", + "issue:47436", + "issue:47448" + ], + "soft_pairs": [ + "issue:44734|issue:46612", + "issue:43577|issue:44554", + "issue:45823|issue:46050", + "issue:43525|issue:43572", + "issue:44964|issue:45405", + "issue:44514|issue:45290", + "issue:43334|issue:43572", + "issue:28218|issue:33357", + "issue:44484|issue:44704", + "issue:43611|issue:46846", + "issue:39692|issue:43295", + "issue:44977|issue:46209", + "issue:43994|issue:46192", + "issue:44488|issue:44843", + "issue:45405|issue:47332", + "issue:42796|issue:44977", + "issue:43257|issue:47404", + "issue:43493|issue:43825", + "issue:43531|issue:47436", + "issue:43572|issue:44291", + "issue:43927|issue:44361", + "issue:43646|issue:44466", + "issue:45491|issue:46002", + "issue:46050|issue:47448", + "issue:44568|issue:45812", + "issue:44493|issue:47448", + "issue:44554|issue:46365", + "issue:44568|issue:46489" + ], + "prepared_review_unit_hash": "ad7ee392d670a5b0e4f047eb086922ec209a47952eb55f5a23b0dc015fc9e964", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36389, + "estimated_input_tokens": 9098, + "estimated_eval_tokens": 18452 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The candidate pairs are mostly semantically similar at a very broad subsystem level, but they do not describe the same concrete bug/change. I would keep them separate; this cluster looks like accidental topical similarity rather than true duplicates.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45812 is the strongest representative because it is a clear, high-signal v4\u2192v5 tokenizer regression with multiple inbound references and a concrete, reproducible failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45812 is the most broadly relevant and well-documented issue in the set, with the clearest regression framing and the most external attention.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44734", + "right": "issue:46612", + "accept": false, + "reason": "Both involve generation cache behavior, but one is a KV-cache continuation crash in serve while the other is beam-search cache reorder being skipped for specific model families; different code paths and fixes." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "One is a dtype propagation bug in BLIP2 loading, the other is an MPS attention correctness issue; unrelated mechanisms and symptoms." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Device placement mismatch during Gemma4 forward and quantized save/load breakage are different problems with different failure points." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Both are missing pad-token config attributes after v5, but they affect different model configs and different attribute names; not the same concrete bug." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "A model-loading failure for Phi-4 multimodal is unrelated to a PEFT version bump that is too new for PyPI." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both touch apply_chat_template, but one fails on batched padding=False input and the other on tool-call assistant messages with no content; distinct edge cases." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Different model families and different missing config attributes; same broad regression class, but not one bug." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Tokenizer spacing after added tokens and a Mac bus error on a CLIP model are unrelated." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "One is a save_pretrained shard-size question, the other is AutoProcessor kwargs not reaching cached_file; no shared underlying defect." + }, + { + "left": "issue:43611", + "right": "issue:46846", + "accept": false, + "reason": "Base-model-prefix loading regression and missing tensor-parallel plan entries are separate model-loading/generation issues." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "Documentation example errors are not the same as a regression in processor.tokenizer access and image passing." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Flash-attention generation failure and FP8 inference failure on B200 are different hardware/model-path issues." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "Wrong SigLIP2 outputs and return_tensors being ignored are separate tokenizer/pipeline behaviors." + }, + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Loading a specific model and offline-mode breakage from unconditional model_info() calls are unrelated." + }, + { + "left": "issue:45405", + "right": "issue:47332", + "accept": false, + "reason": "A dependency-version bump and inability to save an offloaded model are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer BPE marker handling and flash-attention generation failure are different bugs." + }, + { + "left": "issue:43257", + "right": "issue:47404", + "accept": false, + "reason": "Qwen3 MoE weight conversion and device-map loading OOM from conversion transients are not the same change." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 numerical discrepancy and a pipeline error-message regression are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:47436", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE and a missing max_position_embeddings config attribute are separate issues." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Missing pad_token_idx after v5 and an unexpected _is_hf_initialized init argument are different migration regressions." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig save/load token-ID loss and MLukeTokenizer task AttributeError are distinct tokenizer/config bugs." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Custom model initialization breaks in v5, but inconsistent lm_head.weight serialization is a different issue." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both mention NaNs, but the causes and affected components differ completely." + }, + { + "left": "issue:46050", + "right": "issue:47448", + "accept": false, + "reason": "Quantized model save/load breakage and an unrecognized Gemma4 architecture are unrelated." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but one is special-token insertion for mdeberta and the other is wrong token IDs for Granite; not the same concrete bug." + }, + { + "left": "issue:44493", + "right": "issue:47448", + "accept": false, + "reason": "Unexpected position-id keys and unrecognized Gemma4 architecture are different model/config problems." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "An attention correctness issue on MPS is unrelated to an import conflict involving datasets and Trainer." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "Different tokenizer regressions affecting different models and likely different code paths; too broad to merge." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 30, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29925", + "issue:34567", + "issue:37428", + "issue:38175", + "issue:42490", + "issue:42947", + "issue:43278", + "issue:43381", + "issue:43493", + "issue:43504", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43696", + "issue:43701", + "issue:43824", + "issue:43844", + "issue:43873", + "issue:43957", + "issue:43994", + "issue:44112", + "issue:44423", + "issue:44464", + "issue:44493", + "issue:44589", + "issue:44734", + "issue:44849", + "issue:44964", + "issue:45072", + "issue:45084", + "issue:45229", + "issue:45245", + "issue:45357", + "issue:45520", + "issue:45584", + "issue:45823", + "issue:46050", + "issue:46209", + "issue:46285", + "issue:46710", + "issue:46821", + "issue:46832", + "issue:47013", + "issue:47254", + "issue:47404", + "issue:47436" + ], + "soft_pairs": [ + "issue:43278|issue:43701", + "issue:28218|issue:29925", + "issue:43873|issue:47404", + "issue:42490|issue:43696", + "issue:43504|issue:43606", + "issue:43493|issue:43844", + "issue:46821|issue:46832", + "issue:43824|issue:44589", + "issue:44589|issue:45520", + "issue:43278|issue:43844", + "issue:46209|issue:46710", + "issue:43994|issue:44493", + "issue:43606|issue:44112", + "issue:43504|issue:43653", + "issue:44493|issue:46285", + "issue:45072|issue:46832", + "issue:43873|issue:46050", + "issue:44464|issue:45584", + "issue:44423|issue:44734", + "issue:15354|issue:38175", + "issue:43572|issue:47436", + "issue:47013|issue:47254", + "issue:45084|issue:45520", + "issue:45245|issue:46710", + "issue:42947|issue:43381", + "issue:44849|issue:45357", + "issue:45229|issue:45823", + "issue:28218|issue:34567", + "issue:43844|issue:44964", + "issue:44849|issue:46209", + "issue:37428|issue:43957" + ], + "prepared_review_unit_hash": "ba23d0a69162676b807f4e884d5ff0b3fa4712d171aba629bcb766733995ed23", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37082, + "estimated_input_tokens": 9271, + "estimated_eval_tokens": 18798 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All items are issues, not PRs. The cluster is a loose mix of unrelated Transformers bugs; none of the soft pairs look like the same underlying problem, and issue 43873 is the strongest single anchor by activity and breadth.", + "confidence": 0.93, + "canonical_issue_reason": "No true duplicate cluster is evident; issue 43873 is the most active and broadly framed bug report, so it is the best anchor among these issues.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43873 has the highest engagement and a clear reproducible bug description, making it the best overall representative of the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "Different symptoms and code paths: eval dtype drift vs checkpoint resume key mismatch." + }, + { + "left": "issue:28218", + "right": "issue:29925", + "accept": false, + "reason": "Tokenizer spacing after added token is unrelated to SigLIP fast-tokenizer enforcement." + }, + { + "left": "issue:43873", + "right": "issue:47404", + "accept": false, + "reason": "Both involve loading/offloading, but one is quantization offload behavior and the other is transient OOM during conversion." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Import warning on trailing slash is unrelated to a GPT-oss CUDA OOM." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy-field model loading failure is unrelated to CPU offload device mismatch." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "SigLIP2 HF/JAX discrepancy is a model correctness issue, not the ZeRO-3 gradient blowup described here." + }, + { + "left": "issue:46821", + "right": "issue:46832", + "accept": false, + "reason": "Different model bugs: TimesFM window_size attribute error vs RT-DETR feature-level mismatch." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing a missing Qwen class is not the same as a Float8 storage lookup failure." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage resolution and flash_attn import-key handling are separate import/runtime issues." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Eval dtype change and gradient explosion under ZeRO-3 are unrelated failures." + }, + { + "left": "issue:46209", + "right": "issue:46710", + "accept": false, + "reason": "FP8 inference OOM on B200 is not the same as an accuracy regression after upgrade." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 nonsense outputs and unexpected position-id keys are distinct model bugs." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "A real device-mismatch bug is not the same as a stale CI test." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Pretrained-load legacy field error is unrelated to BigBirdTokenizer mask-token registration." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id keys and CLIP text-model structural inconsistency are different regressions." + }, + { + "left": "issue:45072", + "right": "issue:46832", + "accept": false, + "reason": "Bfloat16 dtype mismatch in inference is unrelated to RT-DETR forward shape/length error." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Quantized offloading and save/load breakage for sinq are different quantization bugs." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency is unrelated to Whisper empty-transcription failure." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Continuous-batching multimodal crash and KV-cache continuation indexing are different serve paths." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp limitation is unrelated to SigLIP2 zero-probability outputs." + }, + { + "left": "issue:43572", + "right": "issue:47436", + "accept": false, + "reason": "Missing pad_token_idx and missing max_position_embeddings are different config regressions." + }, + { + "left": "issue:47013", + "right": "issue:47254", + "accept": false, + "reason": "xLSTM train-mode unpacking crash is unrelated to PEFT+ZeRO-3 checkpoint error." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template-node compilation failure is unrelated to flash_attn import lookup on Python 3.13." + }, + { + "left": "issue:45245", + "right": "issue:46710", + "accept": false, + "reason": "Category-count limit error is unrelated to the DeepSeek accuracy regression." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Gradient-checkpointing ineffectiveness in PEFT LoRA is not the same as being unsupported in eval mode." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Qwen3.5 hidden-states bug and incorrect visual-encoder save keys are different regressions." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Multi-GPU CUDA OOM and device mismatch with device_map='auto' are separate Gemma4 issues." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer added-token spacing is unrelated to TrainerState token-count tracking." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "ZeRO-3 gradient anomaly is unrelated to Phi-4 multimodal loading failure." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 hidden-states bug is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "Flash-attention import failure and meta-device model loading breakage are different infrastructure 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": 28, + "review_unit_index": 31, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28370", + "issue:29942", + "issue:30333", + "issue:34689", + "issue:37428", + "issue:41720", + "issue:43065", + "issue:43295", + "issue:43334", + "issue:43404", + "issue:43450", + "issue:43493", + "issue:43525", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43575", + "issue:43701", + "issue:43824", + "issue:43844", + "issue:43873", + "issue:43937", + "issue:44186", + "issue:44554", + "issue:44610", + "issue:44849", + "issue:44869", + "issue:44977", + "issue:45072", + "issue:45198", + "issue:45200", + "issue:45229", + "issue:45406", + "issue:45446", + "issue:45588", + "issue:45759", + "issue:45823", + "issue:46121", + "issue:46192", + "issue:46369", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46821", + "issue:47145", + "issue:47332", + "issue:47404" + ], + "soft_pairs": [ + "issue:45200|issue:45229", + "issue:41720|issue:46710", + "issue:43493|issue:43701", + "issue:47332|issue:47404", + "issue:43937|issue:46701", + "issue:44849|issue:47145", + "issue:43295|issue:43450", + "issue:44977|issue:46701", + "issue:43065|issue:46369", + "issue:43404|issue:46710", + "issue:44554|issue:45446", + "issue:45588|issue:45759", + "issue:45072|issue:45198", + "issue:43295|issue:43873", + "issue:44869|issue:46752", + "issue:43572|issue:46821", + "issue:29942|issue:30333", + "issue:43540|issue:45200", + "issue:45823|issue:46192", + "issue:28370|issue:34689", + "issue:45406|issue:46121", + "issue:43295|issue:44610", + "issue:44186|issue:46192", + "issue:43334|issue:43525", + "issue:43450|issue:43575", + "issue:37428|issue:43824", + "issue:45200|issue:45823", + "issue:43550|issue:43844" + ], + "prepared_review_unit_hash": "f31e91eaba5ad0a2f10fb6933f76b3c36a6dac4fd555563da707417c51650caf", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35518, + "estimated_input_tokens": 8880, + "estimated_eval_tokens": 18016 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This looks like a loose topical cluster of mostly unrelated Transformers issues (model loading, multimodal processors, offloading, flash-attn, configs). I do not see any soft pair that is clearly the same underlying bug/change, so no canonical duplicate stands out.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Gemma4 tokenizer/input-default bug vs multi-GPU inference OOM; different failure modes and code paths." + }, + { + "left": "issue:41720", + "right": "issue:46710", + "accept": false, + "reason": "Qwen3 auto device-map CUDA assert vs DeepSeek accuracy regression after upgrade; unrelated models and symptoms." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 implementation discrepancy vs resume-from-checkpoint key mismatch; no shared concrete bug." + }, + { + "left": "issue:47332", + "right": "issue:47404", + "accept": false, + "reason": "Both involve loading/offloading, but one is save-time failure for a specific model and the other is load-time OOM from conversion transients; not the same bug." + }, + { + "left": "issue:43937", + "right": "issue:46701", + "accept": false, + "reason": "Invalid GenerationConfig vs image-text-to-text pipeline failure; different subsystems and root causes." + }, + { + "left": "issue:44849", + "right": "issue:47145", + "accept": false, + "reason": "Qwen3.5 hidden_states regression vs FP8 weight_scale not applied on load; distinct correctness issues." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression vs batched video input shape bug; separate processor problems." + }, + { + "left": "issue:44977", + "right": "issue:46701", + "accept": false, + "reason": "Flash-attention generation bug vs DiffusionGemma pipeline not working; unrelated." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Sam3PixelDecoder dummy layer vs sam3-litetext memory overflow; different implementation issues." + }, + { + "left": "issue:43404", + "right": "issue:46710", + "accept": false, + "reason": "lm_head weight tying bug vs accuracy regression; not the same code-path." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "MPS attention correctness bug vs flex_attention import version check; different layers entirely." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "flash_attention.py s_aux=None crash vs AutoModelForCausalLM.from_config text_config unwrap issue; unrelated." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "bfloat16 dtype mismatch bugs vs Wav2Vec2 save_pretrained/tokenization failure; not a duplicate." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor.tokenizer regression vs quantized offloading behavior; only broad subsystem overlap." + }, + { + "left": "issue:44869", + "right": "issue:46752", + "accept": false, + "reason": "Whisper timestamp decoding crash vs empty chat-template IndexError; different text-processing paths." + }, + { + "left": "issue:43572", + "right": "issue:46821", + "accept": false, + "reason": "Missing pad_token_idx vs missing window_size attribute; different config fields and models." + }, + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention 2 test failures vs MLflow reporting on failed training jobs; unrelated." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Qwen3OmniMoe video-processing ValueError vs Gemma4 token-type-id defaulting; separate bugs." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Gemma4 device_map auto placement mismatch vs return_tensors being ignored; different behavior and fixes." + }, + { + "left": "issue:28370", + "right": "issue:34689", + "accept": false, + "reason": "Missing vocab_file for custom SentencePiece models vs Llama 3.2 Vision load breakage; different model-loading problems." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer vs rope validation list-type error; no shared underlying bug." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "processor.tokenizer regression vs OmDet processor image-size mismatch; different processor code paths." + }, + { + "left": "issue:44186", + "right": "issue:46192", + "accept": false, + "reason": "LayoutLMv2Tokenizer NER padding/truncation crash vs return_tensors being ignored; unrelated tokenization issues." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Both mention missing pad_token_id, but they affect different model families/config classes and likely need separate fixes." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Batched video processor shape bug vs tensor-parallel OOM on Qwen2 loading; no common bug." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Missing flash-attention support import vs missing Qwen2.5-VL class import; unrelated import errors." + }, + { + "left": "issue:45200", + "right": "issue:45823", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaulting issue vs Gemma4 PLE device mismatch; separate defects." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "torch.compile SDPA failure in Bamba vs DeepSpeed ZeRO-3 gradient anomaly; different training/runtime bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 32, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:27957", + "issue:28218", + "issue:28370", + "issue:29870", + "issue:29942", + "issue:35141", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43335", + "issue:43450", + "issue:43452", + "issue:43493", + "issue:43550", + "issue:43611", + "issue:43653", + "issue:43819", + "issue:43824", + "issue:43856", + "issue:43950", + "issue:43975", + "issue:44112", + "issue:44442", + "issue:44466", + "issue:44521", + "issue:44561", + "issue:44737", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45325", + "issue:45381", + "issue:45520", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46512", + "issue:46894", + "issue:46899", + "issue:47405", + "issue:47448" + ], + "soft_pairs": [ + "issue:45520|issue:46178", + "issue:42915|issue:43824", + "issue:39692|issue:46192", + "issue:27957|issue:28370", + "issue:29870|issue:29942", + "issue:42490|issue:43493", + "issue:43295|issue:43493", + "issue:43824|issue:43975", + "issue:43278|issue:43856", + "issue:46192|issue:47448", + "issue:43975|issue:45520", + "issue:43262|issue:43450", + "issue:44521|issue:45290", + "issue:28218|issue:35141", + "issue:43950|issue:46285", + "issue:43653|issue:44112", + "issue:44737|issue:46894", + "issue:43452|issue:46512", + "issue:45200|issue:45325", + "issue:44561|issue:46234", + "issue:43819|issue:44112", + "issue:44466|issue:45127", + "issue:44442|issue:45198", + "issue:43335|issue:43550", + "issue:45381|issue:46209", + "issue:45200|issue:46899", + "issue:43611|issue:47405" + ], + "prepared_review_unit_hash": "8a655a4bfe9a60298717d729785264c805f3cb70360ff07d06af51345b2fde5e", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35664, + "estimated_input_tokens": 8916, + "estimated_eval_tokens": 18088 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is not a duplicate set: it mixes many unrelated bug reports across tokenization, model loading, multimodal processing, quantization, CI, and architecture-specific regressions. None of the soft-similarity edges look like the same underlying bug/change.", + "confidence": 0.97, + "canonical_issue_reason": "No single canonical issue stands out because the items are heterogeneous and the proposed links are only superficial; they do not share a concrete code path or failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "If a representative issue is needed, issue 43493 is the strongest standalone report: it is specific, reproducible, and describes a clear model-vs-JAX discrepancy.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Different subsystems and failures: flash_attn import key error vs backend registry validation for tensorflow_text." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Unrelated model-specific FP8 inference failure vs an import error for a missing Qwen2.5-VL class." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "Docs/example errors are not the same bug as return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:27957", + "right": "issue:28370", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to missing vocab_file on custom SentencePiece models." + }, + { + "left": "issue:29870", + "right": "issue:29942", + "accept": false, + "reason": "A Bart generate() UnboundLocalError is not the same as failing Flash Attention 2 tests." + }, + { + "left": "issue:42490", + "right": "issue:43493", + "accept": false, + "reason": "Local-path import warning and SigLIP2 implementation discrepancy affect different loading/behavior paths." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "Processor/tokenizer regression in custom model code is unrelated to the SigLIP2 JAX-vs-HF discrepancy." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "ImportError for a vision model class and detokenization for DeepSeek are distinct bugs." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "A dtype regression during evaluate is not the same as Qwen3 MoE memory inefficiency during training." + }, + { + "left": "issue:46192", + "right": "issue:47448", + "accept": false, + "reason": "Ignored return_tensors behavior is unrelated to Gemma4 architecture recognition/loading." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Detokenization output bugs and flash_attn import issues do not share a code path." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate handling is unrelated to batched video processor output shape." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both touch chat templates, but one is multimodal assistant masks and the other is tool-call tokenization crash; different bugs." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Extra space after added token is a tokenizer formatting issue, not the same as embedding reinitialization on resize." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Non-persistent buffer corruption in from_pretrained is unrelated to CLIP text model structure divergence." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and GraniteSpeech CI device override are unrelated issues." + }, + { + "left": "issue:44737", + "right": "issue:46894", + "accept": false, + "reason": "CPU arange device placement in XLNet is not the same as the CohereASR loss double-shift bug." + }, + { + "left": "issue:43452", + "right": "issue:46512", + "accept": false, + "reason": "gguf_file loading failures are unrelated to from_config leaving config.dtype stale." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaults and Qwen2.5-VL rope indexing are different multimodal bugs." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Removing is_torch_fx_available and missing is_falcon_mamba_ssm_available are separate import_utils regressions." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "DAC forward/from_latents mismatch is unrelated to a stale CI test on GraniteSpeech." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Serialization inconsistency for lm_head.weight is not the same as model collapse after LoRA merge with extended vocabulary." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "FastSpeech2ConformerTokenizer loading failure and Wav2Vec2 save_pretrained/tokenization failure are different model-specific problems." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "SwitchTransformers sparse-layer construction and Bamba torch.compile SDPA failure are unrelated code paths." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Qwen2.5-VL video position_id bug is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:45200", + "right": "issue:46899", + "accept": false, + "reason": "Gemma4 token-type-id defaults and 4-bit multimodal dtype casting are different quantization/input bugs." + }, + { + "left": "issue:43611", + "right": "issue:47405", + "accept": false, + "reason": "base_model_prefix loading breakage and silent conversion skip on model_type mismatch are distinct loading regressions." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 33, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28282", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:37428", + "issue:38175", + "issue:41720", + "issue:41950", + "issue:42490", + "issue:43335", + "issue:43425", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43606", + "issue:43653", + "issue:43819", + "issue:43827", + "issue:43854", + "issue:43901", + "issue:44112", + "issue:44292", + "issue:44610", + "issue:44792", + "issue:44843", + "issue:44855", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45200", + "issue:45412", + "issue:45520", + "issue:45563", + "issue:45593", + "issue:45759", + "issue:45923", + "issue:46209", + "issue:46285", + "issue:46431", + "issue:46512", + "issue:46566", + "issue:46821", + "issue:46932", + "issue:47405", + "issue:47427", + "issue:47436", + "issue:47448" + ], + "soft_pairs": [ + "issue:44610|issue:44857", + "issue:37428|issue:43425", + "issue:44898|issue:45412", + "issue:44843|issue:47427", + "issue:45200|issue:46932", + "issue:43819|issue:45083", + "issue:43335|issue:43653", + "issue:41720|issue:46209", + "issue:43606|issue:45083", + "issue:44857|issue:45593", + "issue:44792|issue:45563", + "issue:46431|issue:46566", + "issue:43525|issue:47436", + "issue:41950|issue:47448", + "issue:46512|issue:46821", + "issue:43335|issue:43504", + "issue:45923|issue:47405", + "issue:28282|issue:33357", + "issue:45759|issue:46285", + "issue:43550|issue:44112", + "issue:29942|issue:32090", + "issue:43827|issue:43901", + "issue:44855|issue:45083", + "issue:43335|issue:43606", + "issue:38175|issue:45412", + "issue:44292|issue:45520", + "issue:43335|issue:44112", + "issue:42490|issue:43854" + ], + "prepared_review_unit_hash": "c036f11a1fc5d117ac9fcc2eb4fda95fc44e8d17775160006c49888874e045b7", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35745, + "estimated_input_tokens": 8937, + "estimated_eval_tokens": 18130 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated issues that only look similar at the title/token level. None of the soft-edge pairs appears to be the same underlying bug or change, so all soft candidates should be rejected. I picked issue 43335 as the canonical issue only because it is the most connected node in the similarity graph, not because it truly represents a duplicate group.", + "confidence": 0.68, + "canonical_issue_reason": "issue:43335 has the highest soft-edge connectivity in this cluster and is a concrete, well-scoped model bug, but it is only an arbitrary representative because the cluster is not actually a coherent duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43335 is the best available representative by graph centrality and specificity; however, the surrounding items are too heterogeneous for a strong canonical choice.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different bugs in different code paths: OmDet image-size preprocessing mismatch vs LwDetr float16/CUDA loss crash." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "An import error from a missing internal symbol is not the same as a Torch 2.10 compatibility issue." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Perceiver non-default resolution handling and RT-DETR model memory release are unrelated model-specific problems." + }, + { + "left": "issue:44843", + "right": "issue:47427", + "accept": false, + "reason": "Offline tokenizer patching via hub access is unrelated to redundant initialization checks on non-CUDA hardware." + }, + { + "left": "issue:45200", + "right": "issue:46932", + "accept": false, + "reason": "Both mention Gemma 4, but one is about missing text-only token type defaults and the other about quantized embedder dtype casting." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "DAC latent reconstruction mismatch and qwen3_omni_moe feature-length helper behavior are different audio-model bugs." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "Sparse-layer construction in SwitchTransformers is unrelated to BigBird tokenizer special-token registration." + }, + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Same broad Qwen family, but different failures: auto device mapping cuda assert vs FP8 inference on B200 hardware." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "CPU offload device mismatch in Bark is unrelated to qwen3_omni_moe output-length calculation." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different model/components and different failure modes: AMP crash vs missing auxiliary losses." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "janus image-generation test failure and a stale generation warning are not the same issue." + }, + { + "left": "issue:46431", + "right": "issue:46566", + "accept": false, + "reason": "Chat-template availability in a processor and weights stuck on meta under device_map='auto' are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:47436", + "accept": false, + "reason": "Both are missing-config-attribute reports, but they affect different models and different attributes." + }, + { + "left": "issue:41950", + "right": "issue:47448", + "accept": false, + "reason": "Video pipeline processor selection and Gemma4 architecture recognition are unrelated subsystems." + }, + { + "left": "issue:46512", + "right": "issue:46821", + "accept": false, + "reason": "A stale dtype field after from_config and a TimesFM window_size AttributeError are unrelated config bugs." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and Beit pretrained preset loading are different model-specific defects." + }, + { + "left": "issue:45923", + "right": "issue:47405", + "accept": false, + "reason": "Missing gradient-checkpointing metadata on a trust_remote_code model is unrelated to silent conversion skipping under the wrong Auto class." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "A PyTorch-missing import error and a MacOS bus error in CLIP loading are different environment/runtime failures." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Both involve composite-model structure, but one is AutoModelForCausalLM text_config unwrapping and the other is CLIP class structural inconsistency." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and GraniteSpeech stale device override are unrelated model/runtime bugs." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 test failures and NoneType GPU broadcast errors in Trainer do not share the same root cause." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Docs for pipeline removals and text-classification return_all_scores docs drift are separate documentation issues." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "A Python 3.13 torch.jit parsing failure is unrelated to qwen3_omni_moe feature-length behavior." + }, + { + "left": "issue:43335", + "right": "issue:43606", + "accept": false, + "reason": "SwitchTransformers sparse-layer logic and Bark CPU offload device mismatch are unrelated bugs." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "Zero-probabilities behavior in a SigLIP2 model and RT-DETR memory retention are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:45520", + "accept": false, + "reason": "Qwen-3-8B-NVFP4 runtime failure and flash_attn import key errors are different problems at different layers." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and GraniteSpeech stale device-override CI failure are unrelated." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "A local-path trailing-slash warning and a model-loading unit-test failure for GLM-4.7-Flash are not the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 34, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:25251", + "issue:29127", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:42175", + "issue:42490", + "issue:43278", + "issue:43329", + "issue:43335", + "issue:43388", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43701", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44448", + "issue:44493", + "issue:44514", + "issue:44521", + "issue:44617", + "issue:44743", + "issue:44779", + "issue:44857", + "issue:45072", + "issue:45083", + "issue:45290", + "issue:45584", + "issue:45588", + "issue:45823", + "issue:45859", + "issue:46326", + "issue:46424", + "issue:46697", + "issue:46858" + ], + "soft_pairs": [ + "issue:44617|issue:46697", + "issue:46424|issue:46858", + "issue:44206|issue:45584", + "issue:43819|issue:44493", + "issue:43388|issue:43526", + "issue:45588|issue:45859", + "issue:43335|issue:45083", + "issue:10000|issue:15354", + "issue:43606|issue:44857", + "issue:39692|issue:44493", + "issue:45290|issue:46326", + "issue:37428|issue:44351", + "issue:44206|issue:45823", + "issue:39692|issue:43526", + "issue:44448|issue:44743", + "issue:34689|issue:35141", + "issue:44514|issue:44521", + "issue:43720|issue:44112", + "issue:44857|issue:45072", + "issue:43701|issue:45083", + "issue:29127|issue:32090", + "issue:44448|issue:44779", + "issue:25251|issue:32090", + "issue:43550|issue:45083", + "issue:32090|issue:33357", + "issue:37428|issue:42175", + "issue:42490|issue:43575", + "issue:43278|issue:43825", + "issue:29942|issue:35141", + "issue:43278|issue:43329" + ], + "prepared_review_unit_hash": "8149e8fbede4b4298dc508751171ce55e508f68efeb093bfea694fe0d98bd53b", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36395, + "estimated_input_tokens": 9099, + "estimated_eval_tokens": 18454 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is thematically broad but not duplicate-heavy: it mixes unrelated Transformers bugs across chat templates, caching, dtype/device mismatches, model loading, preprocessing, and CI/test failures. None of the soft-similarity pairs look like the same underlying bug or change.", + "confidence": 0.95, + "canonical_issue_reason": "Issue 44521 is the best representative anchor because it is open, well-described, has inbound references, and concerns a concrete multimodal chat-template bug with clear reproduction details.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44521 is the strongest global issue candidate: it has the clearest user-facing failure, active discussion, and a specific code path (`apply_chat_template` for multimodal inputs).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44617", + "right": "issue:46697", + "accept": false, + "reason": "One is a video-model CUDA OOM, the other is an image-processor padding bug; different subsystems and failure modes." + }, + { + "left": "issue:46424", + "right": "issue:46858", + "accept": false, + "reason": "Both involve cache behavior, but one is a config limit bug and the other is a device-tensor arithmetic issue in `_prefill`." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Different audio-related problems: unsupported feature-extractor arg vs empty-transcription generation failure." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "Different code paths: DAC latent conversion mismatch vs position-id key warnings in model loading." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Both mention labels, but one is distributed metric gathering and the other is BEiT label reduction; not the same bug." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Both touch model internals, but one is a flash-attention helper crash on `s_aux=None` and the other is CPU-offload incompatibility in rotary embeddings." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "Unrelated bugs in different model components: SwitchTransformers config creation vs qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "Different models and failures: DistilBERT issue vs TorchScript GeneratorExp export incompatibility." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "Both are device/dtype-related, but they affect different models and distinct mechanisms (CPU offload mismatch vs AMP/CUDA loss crash)." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "A documentation/example bug plus quantization failure is not the same as a model position-id warning regression." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both involve chat templating, but one crashes on assistant tool-call messages and the other breaks `continue_final_message` for mistral_common tokenizers." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different missing symbols and different modules; not the same underlying break." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Different models and issues: unsupported feature-extractor arg vs Gemma4 device-mapping mismatch during forward." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "Both are vision/data-processing related, but one is a SigLIP2 docs example and the other is BEiT fast processor label handling." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Both are v5 behavioral regressions, but one concerns Pegasus output changes while the other is recurrent-state reset logic in qwen3 modular code." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Different problems: Llama 3.2 Vision loading regression vs token-embedding resize reinitialization after `post_init`." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Both are chat-template issues, but one crashes on batched padded input and the other returns all-zero assistant masks; distinct bugs." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "A BitNet loading bug and a GraniteSpeech CI test flake are unrelated." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both involve dtype/device mismatches, but in different models and execution settings; not the same code-path defect." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch and qwen3_omni_moe feature-length behavior are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "Different failure classes: clearer error messaging for layoutlmv3 vs NoneType broadcast error in Trainer." + }, + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both are output regressions, but one is Pegasus generation drift and the other is tokenizer corruption in Deepseek." + }, + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to `_gpu_broadcast_one` NoneType failures." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Trainer broadcast NoneType error and MacOS bus error in community CLIP are distinct issues." + }, + { + "left": "issue:37428", + "right": "issue:42175", + "accept": false, + "reason": "An import-symbol regression is unrelated to missing TensorFlow in a pip extras install." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "Local-path trailing-slash warning and tensor-parallel OOM are different problems." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "BF16/FP32 dtype drift during eval is unrelated to a pipeline error-message regression." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures do not match an embedding resize reinitialization bug." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "A dtype change bug and an undefined-variable multimodal token-count bug 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": 28, + "review_unit_index": 35, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:26413", + "issue:27957", + "issue:29870", + "issue:29925", + "issue:29942", + "issue:30064", + "issue:31501", + "issue:32090", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:39692", + "issue:40444", + "issue:42490", + "issue:43257", + "issue:43278", + "issue:43299", + "issue:43381", + "issue:43526", + "issue:43873", + "issue:43927", + "issue:44360", + "issue:44448", + "issue:44464", + "issue:44530", + "issue:44561", + "issue:44589", + "issue:44855", + "issue:44977", + "issue:45127", + "issue:45229", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", + "issue:45593", + "issue:45698", + "issue:45920", + "issue:46192", + "issue:46234", + "issue:46399", + "issue:46424", + "issue:46655", + "issue:47000", + "issue:47276", + "issue:47436" + ], + "soft_pairs": [ + "issue:45310|issue:45478", + "issue:45593|issue:47000", + "issue:45459|issue:46234", + "issue:27957|issue:34689", + "issue:26413|issue:31501", + "issue:45290|issue:45381", + "issue:45127|issue:45245", + "issue:40444|issue:43526", + "issue:29870|issue:32090", + "issue:44360|issue:44561", + "issue:44530|issue:47276", + "issue:44448|issue:44977", + "issue:44464|issue:46424", + "issue:32090|issue:33290", + "issue:44589|issue:47436", + "issue:44855|issue:46399", + "issue:29870|issue:33290", + "issue:42490|issue:43873", + "issue:43278|issue:43381", + "issue:44977|issue:45229", + "issue:45290|issue:46655", + "issue:45920|issue:46192", + "issue:43257|issue:43299", + "issue:29925|issue:34689", + "issue:43927|issue:45698", + "issue:29942|issue:30064", + "issue:33357|issue:39692", + "issue:43873|issue:45229" + ], + "prepared_review_unit_hash": "24c001cc9d1e68a22c8883ec146a88211dd60858296f3fda52ee3864537eb6ab", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36028, + "estimated_input_tokens": 9007, + "estimated_eval_tokens": 18270 + }, + "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 Qwen3.5 MoE from_pretrained regression reports. The rest are related only at a broad subsystem level (training, generation, tokenizer, multimodal, deepspeed, etc.) and should stay separate.", + "confidence": 0.96, + "canonical_issue_reason": "issue:45310 is the better canonical: it was opened first and has the fuller report, while issue:45478 is essentially the same Qwen3.5 MoE from_pretrained failure with a near-identical title.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the strongest representative issue in the only true duplicate pair; it most cleanly captures the underlying Qwen3.5 MoE loading regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Near-identical titles and the same concrete failure: transformers>=5.4.0 Qwen3.5 MoE from_pretrained error. Same underlying bug." + }, + { + "left": "issue:45593", + "right": "issue:47000", + "accept": false, + "reason": "Both concern detection losses, but one is auxiliary-loss handling in D-FINE and the other is an infeasible cost matrix in RTDetrLoss; different bugs and code paths." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Both touch import-utils/tokenizer plumbing, but one is protobuf decode error masking and the other is a missing availability helper; unrelated issues." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Both are model-loading/runtime problems, but for different models and different failures. Too broad to treat as duplicates." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Both affect training startup, but one is checkpoint resume/epoch iterator handling and the other is a FineWeb SLM training launch failure; not the same bug." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "One is chat template tokenization with tool-call assistant messages; the other is Qwen2.5-VL video vision position ids. Different functionality." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA merging with tied embeddings is unrelated to the category-count limit error." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Both involve vision processors, but one is iterable multi-image VL finetuning and the other is BeitImageProcessorFast label reduction; different code paths." + }, + { + "left": "issue:29870", + "right": "issue:32090", + "accept": false, + "reason": "Bart generate() decoder mask initialization and Trainer _gpu_broadcast_one NoneType errors are separate bugs." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "A missing ReLU in the DSA indexer is unrelated to removal of is_torch_fx_available breaking trust_remote_code models." + }, + { + "left": "issue:44530", + "right": "issue:47276", + "accept": false, + "reason": "PagedAttentionCache group-type failure and a misleading synced_gpus warning are not the same underlying issue." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different model regressions: Pegasus v4/v5 output changes versus Qwen3.5 flash-attention generation problems." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation with compiled forward and cache max length handling are distinct generation/cache behaviors." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "A NoneType broadcast failure in Trainer is not the same as an Adafactor+DeepSpeed OOM." + }, + { + "left": "issue:44589", + "right": "issue:47436", + "accept": false, + "reason": "Float8 storage lookup failure and missing max_position_embeddings on Qwen2_5OmniConfig are unrelated." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Both mention decorators/registration, but one is a Python 3.13 parsing/indentation issue and the other is a Qwen3.5 kernel registration crash." + }, + { + "left": "issue:29870", + "right": "issue:33290", + "accept": false, + "reason": "Bart generation mask bug and DeepSpeed Adafactor OOM are not duplicates." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Local-path import warning with trailing slash and quantization offloading problems are different concerns." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "One is an eval-time dtype change for embeddings; the other is a restriction on using gradient checkpointing in eval mode. Related but not the same bug." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Qwen3.5 flash-attention generation and Gemma4 multi-GPU OOM are unrelated failures." + }, + { + "left": "issue:45290", + "right": "issue:46655", + "accept": false, + "reason": "Chat-template tokenization with tool calls and SmolVLM video padding are different multimodal bugs." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Wrong token IDs in AutoTokenizer and silent return_tensors ignoring are separate tokenizer API issues." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Both involve Qwen3 MoE loading, but one is accelerate+deepspeed weight conversion and the other is a broader from_pretrained failure for Qwen3VL MoE; not clearly the same bug." + }, + { + "left": "issue:29925", + "right": "issue:34689", + "accept": false, + "reason": "SigLIP tokenizer fast-mode enforcement and Llama 3.2 Vision model loading break are unrelated." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Config token-ID persistence and wrong custom-module loading after save_pretrained are both serialization-related, but they are different failure modes." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Failing Flash Attention 2 tests and void segmentation-map processing are separate problems." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error on community CLIP loading and SigLIP2 documentation errors are different issues." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Quantization/offloading behavior and Gemma4 multi-GPU OOM are not the same code-path problem." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The conclusions are grounded in the packet and the duplicate call is conservative: issue:45310 and issue:45478 have near-identical titles and clearly point to the same Qwen3.5 MoE from_pretrained regression. The remaining pairs are rejected with appropriately cautious, title-level distinctions rather than overstated duplicate claims." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 36, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:23647", + "issue:29127", + "issue:29870", + "issue:34567", + "issue:34689", + "issue:41720", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43278", + "issue:43334", + "issue:43381", + "issue:43412", + "issue:43531", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43756", + "issue:43976", + "issue:43994", + "issue:44077", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44361", + "issue:44387", + "issue:44442", + "issue:44448", + "issue:44466", + "issue:44479", + "issue:44561", + "issue:44743", + "issue:44779", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45229", + "issue:45440", + "issue:45563", + "issue:45588", + "issue:45803", + "issue:45823", + "issue:46121", + "issue:46421", + "issue:46858", + "issue:47276" + ], + "soft_pairs": [ + "issue:45588|issue:45823", + "issue:44186|issue:45198", + "issue:44561|issue:44779", + "issue:44448|issue:44466", + "issue:43412|issue:45588", + "issue:44871|issue:44977", + "issue:23647|issue:29870", + "issue:43381|issue:43701", + "issue:46421|issue:46858", + "issue:44361|issue:45198", + "issue:29127|issue:34567", + "issue:41720|issue:45229", + "issue:43278|issue:43575", + "issue:43334|issue:43531", + "issue:44265|issue:45198", + "issue:44206|issue:44479", + "issue:44977|issue:45563", + "issue:43653|issue:44442", + "issue:43611|issue:46121", + "issue:42915|issue:42994", + "issue:43976|issue:45440", + "issue:45198|issue:45803", + "issue:43065|issue:43994", + "issue:29870|issue:34689", + "issue:44077|issue:47276", + "issue:43653|issue:44743", + "issue:43756|issue:45440", + "issue:43575|issue:44387" + ], + "prepared_review_unit_hash": "cd2e5c61b45e07f9831539742959e03d8b1a2f93cf2c3ea6b36292ab159f3a07", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35569, + "estimated_input_tokens": 8893, + "estimated_eval_tokens": 18042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated tokenizer, generation, quantization, cache, export, and training/runtime issues across different models and code paths. None of the soft pairs look like the same underlying bug/change, so no duplicate merge is warranted.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out because the cluster is not a coherent duplicate set; the items span distinct subsystems and failure modes.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue candidate is appropriate here because these reports are not variants of one problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Both are forward/inference bugs, but one is a flash-attention AttributeError and the other is a device-map mismatch in Gemma4; different code paths and fixes." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizers, different models, and different failure modes (NER/padding crash vs save_pretrained/tokenization failure)." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "One is a v5 API removal breaking trust_remote_code models; the other is a DeepSeek tokenizer regression. Not the same bug." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Both are v5 regressions, but one changes Pegasus generation output while the other is about lm_head.weight serialization; distinct underlying issues." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Export-time self-training-guard/erfinv problems versus a runtime flash-attention AttributeError; not mergeable as one fix." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id config mismatch is unrelated to Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:23647", + "right": "issue:29870", + "accept": false, + "reason": "WandB sweep custom objective integration and BART generate decoder_attention_mask initialization are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction and resume_from_checkpoint key mismatch are separate trainer issues." + }, + { + "left": "issue:46421", + "right": "issue:46858", + "accept": false, + "reason": "Both touch cache/compile behavior, but one is lazy-init recompilation and the other is a device-tensor seq_length arithmetic bug; different concrete defects." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "Different tokenizer/model families and different symptoms; no shared code-path bug is evident." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "A LayoutLMv3 error-message clarity issue is unrelated to TrainerState token counting not updating." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Both involve large-model inference resource problems, but cudaErrorAssert on Qwen3 is not the same bug as Gemma4 multi-GPU OOM." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "Embedding dtype drift during evaluate and Qwen2 tensor-parallel OOM are unrelated." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "A pad_token_id attribute error in Qwen3-VL is not the same issue as sliding_window handling in Qwen3-MoE." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export failures with torch_compilable_check are unrelated to Wav2Vec2 save/tokenization failures." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Both are regressions, but one is a feature extractor arg mismatch and the other is a video-input regression in Qwen VL models; different fixes." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation failure and a stale paged-generate warning are different problems in generation code." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer special-token registration and FastSpeech2Conformer tokenizer loading are distinct tokenizer issues." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "Model loading with base_model_prefix and rope-parameter dict conversion are unrelated configuration/validation bugs." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "FineGrainedFP8Config failure and quantized model saving failure are different quantization-related workflows, not the same defect." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "A Python version compatibility regression and a DeepSeekV3MoE implementation divergence are unrelated." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 tokenization/save failures and a bare-except in FuyuBatchFeature are separate issues." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "A dummy Conv2d in Sam3PixelDecoder is unrelated to SigLIP2 producing nonsensical outputs through AutoModel/pipeline." + }, + { + "left": "issue:29870", + "right": "issue:34689", + "accept": false, + "reason": "BART generate UnboundLocalError and Llama 3.2 Vision model loading breakage are different code paths." + }, + { + "left": "issue:44077", + "right": "issue:47276", + "accept": false, + "reason": "Optional post_init validation for patchtsmixer is unrelated to the synced_gpus warning text in continuous batching." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "Tokenizer special-token registration and Qwen3.5 cache state reset behavior are unrelated." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and DeepSeekV3MoE divergence are different model implementation bugs." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Tensor-parallel OOM for Qwen2 is not the same as int4 quantization increasing CUDA reserved memory in Transformers 5.x." + } + ] + }, + "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": 28, + "review_unit_index": 37, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:23647", + "issue:27957", + "issue:28282", + "issue:29127", + "issue:33290", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:42490", + "issue:42915", + "issue:43065", + "issue:43381", + "issue:43421", + "issue:43526", + "issue:43575", + "issue:43701", + "issue:43824", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44186", + "issue:44279", + "issue:44442", + "issue:44568", + "issue:44610", + "issue:44625", + "issue:44857", + "issue:45072", + "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", + "issue:47217" + ], + "soft_pairs": [ + "issue:44568|issue:45491", + "issue:43575|issue:43701", + "issue:42490|issue:43381", + "issue:29127|issue:33357", + "issue:43994|issue:45412", + "issue:45491|issue:46693", + "issue:45372|issue:45588", + "issue:44442|issue:45072", + "issue:45406|issue:46701", + "issue:43065|issue:45412", + "issue:44625|issue:45584", + "issue:45250|issue:46640", + "issue:43526|issue:47217", + "issue:43526|issue:45834", + "issue:29127|issue:35141", + "issue:45440|issue:45593", + "issue:45459|issue:45712", + "issue:45372|issue:45406", + "issue:43994|issue:46369", + "issue:43526|issue:44186", + "issue:43824|issue:44279", + "issue:43381|issue:43873", + "issue:43421|issue:45491", + "issue:45372|issue:45459", + "issue:42915|issue:43931", + "issue:33290|issue:35141", + "issue:33290|issue:34689", + "issue:45245|issue:45563", + "issue:27957|issue:33290", + "issue:44857|issue:45588", + "issue:23647|issue:28282", + "issue:44610|issue:45412" + ], + "prepared_review_unit_hash": "a9f0039276d15d1dec458a78b3788e14cbc0dcf7f14910ca518ec7b13286c152", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36484, + "estimated_input_tokens": 9121, + "estimated_eval_tokens": 18498 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair is a safe duplicate: these are mostly unrelated bug reports, regressions, and feature requests across different models and code paths. Best standalone issue is the FA2 long-context regression because it is specific, actionable, and clearly scoped.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:46693 is the strongest standalone issue: it describes a concrete Flash Attention 2 performance regression with long-context generation, a clear repro surface, and actionable scope.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Different failures and models: tokenizer BOS/EOS behavior vs Gemma3 NaN embeddings in mixed-length batching." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "One is TP memory/OOM on Qwen2-57B; the other is a checkpoint key mismatch during resume. Not the same bug." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Import warning for local model paths is unrelated to gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:29127", + "right": "issue:33357", + "accept": false, + "reason": "LayoutLMv3 box validation messaging and a CLIP bus error on macOS are different issues and code paths." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 bad outputs and RT-DETR memory not being released are unrelated model/runtime problems." + }, + { + "left": "issue:45491", + "right": "issue:46693", + "accept": false, + "reason": "Both mention FA2-adjacent context, but one is NaN embeddings and the other is a performance regression; different underlying bugs." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Gemma 4 import failure from mistral_common is not the same as a flash_attention s_aux=None AttributeError." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "AutoTokenizer loading failure and bfloat16 dtype mismatches are distinct problems." + }, + { + "left": "issue:45406", + "right": "issue:46701", + "accept": false, + "reason": "Gemma4Processor serve crash from missing _tokenizer is separate from DiffusionGemma pipeline behavior." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to RT-DETR memory leaks on deletion." + }, + { + "left": "issue:44625", + "right": "issue:45584", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is unrelated to Whisper failing on empty transcription after special-token alignment." + }, + { + "left": "issue:45250", + "right": "issue:46640", + "accept": false, + "reason": "Both are generic Flash Attention 2 topics, but neither identifies the same concrete bug or fixable code path." + }, + { + "left": "issue:43526", + "right": "issue:47217", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels bug and VL image-placeholder failures are different subsystems and symptoms." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "Reduce_labels returning one label and Kosmos2.5 OCR index errors are unrelated." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "LayoutLMv3 error messaging and token-embedding reinitialization on resize are not the same defect." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepseekV3MoE divergence from remote implementation is unrelated to D-FINE auxiliary loss handling." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Hidden tokenizer errors without protobuf and leftover dummy classes are separate maintenance/behavior issues." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "One is an import-time dependency failure; the other is a runtime serve crash with a missing attribute." + }, + { + "left": "issue:43994", + "right": "issue:46369", + "accept": false, + "reason": "Different models and failures: nonsensical SigLIP2 outputs vs Sam3-litetext memory overflow." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "BeitImageProcessorFast reduce_labels and LayoutLMv2Tokenizer NER/padding crashes are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Specific Qwen2_5_VL import error vs a vague dependency issue; not the same underlying bug." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and quantization offloading behavior are different code paths." + }, + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Runtime post-processor updates for special tokens do not match Gemma3 NaN embeddings." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma4 processor import failure and protobuf-related tokenizer error masking are different problems." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and Qwen3-VL-30B weight-shape mismatch are distinct model-loading issues." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "DeepSpeed Adafactor OOM is unrelated to embedding reinitialization after resizing." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "Optimizer OOM and Llama 3.2 Vision model loading breakage are different defects." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Category-cardinality runtime error and a stale generate warning are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta is not the same bug as Adafactor+DeepSpeed OOM." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetrImageLoss AMP CUDA crash and flash_attention s_aux=None AttributeError are separate failures." + }, + { + "left": "issue:23647", + "right": "issue:28282", + "accept": false, + "reason": "Wandb sweeps custom objective support and missing PyTorch import errors do not share an underlying code-path." + }, + { + "left": "issue:44610", + "right": "issue:45412", + "accept": false, + "reason": "OmDet-Turbo processor image-size mismatch and RT-DETR memory retention are unrelated issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 38, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:28282", + "issue:29127", + "issue:29870", + "issue:29925", + "issue:42175", + "issue:42490", + "issue:42617", + "issue:43066", + "issue:43262", + "issue:43316", + "issue:43493", + "issue:43575", + "issue:43643", + "issue:43646", + "issue:43720", + "issue:43749", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44077", + "issue:44206", + "issue:44442", + "issue:44462", + "issue:44479", + "issue:44514", + "issue:44743", + "issue:44811", + "issue:44871", + "issue:44898", + "issue:44913", + "issue:44977", + "issue:45245", + "issue:45381", + "issue:45588", + "issue:45759", + "issue:45859", + "issue:46121", + "issue:46234", + "issue:46399", + "issue:46821", + "issue:47276", + "issue:47312", + "issue:47436", + "issue:47448" + ], + "soft_pairs": [ + "issue:43066|issue:44462", + "issue:45588|issue:46821", + "issue:29870|issue:29925", + "issue:25251|issue:28282", + "issue:44442|issue:44898", + "issue:27957|issue:29127", + "issue:44514|issue:45381", + "issue:46121|issue:46399", + "issue:43761|issue:45859", + "issue:43646|issue:44077", + "issue:46821|issue:47436", + "issue:43316|issue:47312", + "issue:43720|issue:44442", + "issue:44743|issue:44811", + "issue:42175|issue:46234", + "issue:43262|issue:44206", + "issue:44077|issue:44479", + "issue:43873|issue:44871", + "issue:43316|issue:45759", + "issue:42490|issue:44743", + "issue:44977|issue:45245", + "issue:47276|issue:47448", + "issue:43493|issue:43827", + "issue:43749|issue:43931", + "issue:42617|issue:43575", + "issue:43643|issue:44913" + ], + "prepared_review_unit_hash": "ae71bdb3138621f6c338e99afd76e75f3e5b29de5a16850d1c6e3f776d473b88", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34865, + "estimated_input_tokens": 8717, + "estimated_eval_tokens": 17690 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is very noisy: the items span unrelated bugs in tokenizers, model loading, vision/audio processors, configs, docs, and backend availability. None of the soft pairs look like true duplicates or the same fixable code-path issue.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43873 is the broadest, highest-engagement runtime bug in the set, so it is the best single representative even though the overall cluster is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 has the most discussion/inbound refs and the widest apparent user impact, making it the most suitable anchor issue for this noisy group.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Both concern tokenizer behavior, but one is a decoder-type regression and the other is repository tokenizer file selection; different bugs and code paths." + }, + { + "left": "issue:45588", + "right": "issue:46821", + "accept": false, + "reason": "Unrelated failures: flash_attention sink handling vs TimesFM window_size attribute access." + }, + { + "left": "issue:29870", + "right": "issue:29925", + "accept": false, + "reason": "Bart generate() decoder_attention_mask bug and SigLIP fast-tokenizer enforcement are separate tokenizer/generation issues." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to missing PyTorch ImportError during AutoModel loading." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Different model families and failures: FastSpeech2Conformer tokenizer load vs Perceiver image-size interpolation." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to LayoutLMv3 box validation messaging." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both are Qwen2.5-VL, but one is batched chat-template padding and the other is video vision_position_ids; not the same bug." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "Rope-parameter JSON/list handling is unrelated to Qwen3.5 kernel registration/decorator failure." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "CLIP hidden_states regression and Qwen3.5 MoE rotary embedding CPU offload incompatibility are distinct issues." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Generic custom model init breakage is not the same as patchtsmixer's post_init API constraint." + }, + { + "left": "issue:46821", + "right": "issue:47436", + "accept": false, + "reason": "TimesFM window_size attribute error and Qwen2_5OmniConfig missing max_position_embeddings are unrelated config/loading bugs." + }, + { + "left": "issue:43316", + "right": "issue:47312", + "accept": false, + "reason": "Gemma3TextConfig API mismatch is unrelated to kernels 0.16 incompatibility." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet accelerate-loading unpacking bug and FastSpeech2Conformer tokenizer load failure are unrelated." + }, + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Qwen recurrent-state reset bug and Whisper batch_decode skip_special_tokens behavior affect different code paths." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Backend packaging issue with pip extras is unrelated to a missing import_utils availability helper." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting and LasrFeatureExtractor center-arg crash are separate audio bugs." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "PatchTSMixer post_init restriction is unrelated to Qwen video-input regression." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization/offloading failure and Gemma eos_token_id mismatch are different runtime/config issues." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Both involve model config APIs, but Gemma3TextConfig discrepancy and Qwen composite text_config unwrapping are different defects." + }, + { + "left": "issue:42490", + "right": "issue:44743", + "accept": false, + "reason": "Local-path import warning has no relation to Qwen recurrent-state reset behavior." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Flash-attention generation failure and category-count limit error are unrelated." + }, + { + "left": "issue:47276", + "right": "issue:47448", + "accept": false, + "reason": "Misleading synced_gpus warning and Gemma4 arch recognition are different issues." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and docs still referencing pipeline() are unrelated." + }, + { + "left": "issue:43749", + "right": "issue:43931", + "accept": false, + "reason": "FSDP CPU-efficient loading breakage and Qwen3-VL weight-shape mismatch are different loading problems." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "3d_parallel.py runtime failure and Qwen2-57B-A14B-Instruct TP OOM are not the same bug." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "Missing fields from trust_remote_code AutoConfig and GPTNeoX rotary_pct reset on reload are separate config persistence bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 39, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:23647", + "issue:27957", + "issue:28218", + "issue:29870", + "issue:32090", + "issue:43232", + "issue:43295", + "issue:43316", + "issue:43388", + "issue:43525", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43761", + "issue:43825", + "issue:43931", + "issue:43950", + "issue:44062", + "issue:44077", + "issue:44079", + "issue:44188", + "issue:44190", + "issue:44442", + "issue:44496", + "issue:44521", + "issue:44534", + "issue:44855", + "issue:44871", + "issue:44964", + "issue:44987", + "issue:45137", + "issue:45216", + "issue:45440", + "issue:45584", + "issue:45803", + "issue:45907", + "issue:46178", + "issue:46506", + "issue:46829", + "issue:46932", + "issue:47217", + "issue:47427", + "issue:47448" + ], + "soft_pairs": [ + "issue:44987|issue:46178", + "issue:43550|issue:43701", + "issue:43316|issue:43525", + "issue:43950|issue:47427", + "issue:27957|issue:29870", + "issue:44964|issue:45440", + "issue:43606|issue:43701", + "issue:43575|issue:43742", + "issue:43825|issue:47448", + "issue:43761|issue:44079", + "issue:44062|issue:45907", + "issue:43316|issue:43716", + "issue:44855|issue:45137", + "issue:43701|issue:46506", + "issue:46829|issue:46932", + "issue:44190|issue:44496", + "issue:43931|issue:45216", + "issue:44442|issue:45803", + "issue:27957|issue:32090", + "issue:23647|issue:28218", + "issue:43295|issue:44871", + "issue:44871|issue:45584", + "issue:43761|issue:44188", + "issue:43232|issue:43388", + "issue:44077|issue:44534", + "issue:16998|issue:22355", + "issue:44521|issue:47217" + ], + "prepared_review_unit_hash": "d78df4f152a4036e683f7a4807ae571210729a86599a55a69653b574fad3f335", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35850, + "estimated_input_tokens": 8963, + "estimated_eval_tokens": 18182 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The soft pairs are mostly false positives: they share broad Transformer themes but describe different code paths, components, or regression types. This is not a coherent duplicate cluster.", + "confidence": 0.96, + "canonical_issue_reason": "No solid canonical issue: the set mixes unrelated bugs across tokenizers, configs, generation, loading, training, and model export/runtime behavior.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44521 is the most active/open item with the most inbound references, so it is the best available cluster anchor, but it still does not unify the rest of the issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "Different failures: model loading for a specific repo vs backend registration for tensorflow_text." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "Unrelated problems: torch.compile SDPA bug vs checkpoint key mismatch on resume." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Both are config/API issues, but one is Gemma3TextConfig discrepancy and the other is missing pad_token_id on Llama4Config." + }, + { + "left": "issue:43950", + "right": "issue:47427", + "accept": false, + "reason": "Different code paths: corrupted non-persistent buffers on load vs redundant HF init checks on non-CUDA hardware." + }, + { + "left": "issue:27957", + "right": "issue:29870", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Bart generate() using decoder_attention_mask before assignment." + }, + { + "left": "issue:44964", + "right": "issue:45440", + "accept": false, + "reason": "Different model families and bugs: Phi-4 multimodal load failure vs DeepSeekV3MoE implementation divergence." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Qwen2 tensor-parallel OOM and MobileLLM config key error are different loading issues." + }, + { + "left": "issue:43825", + "right": "issue:47448", + "accept": false, + "reason": "One is an error-message regression in pipeline(), the other is gemma4 architecture recognition." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "CLIPVisionModel hidden_states output bug is different from ModelOutput key assignment when previous value is None." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer AddedToken argument collision is unrelated to list_repo_files() not forwarding revision." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and Mistral-3 vision dtype mismatch are separate regressions." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 indentation/import parsing issue is unrelated to DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:43701", + "right": "issue:46506", + "accept": false, + "reason": "Checkpoint resume key mismatch and CPU/CUDA output divergence are different problems." + }, + { + "left": "issue:46829", + "right": "issue:46932", + "accept": false, + "reason": "VideoPrism hidden_states misuse and Gemma4 quantized embedder dtype casting do not look like the same bug." + }, + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading in a training script is unrelated to unrecognized model/config loading." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "One is a shape mismatch when loading a model, the other is a save_pretrained regression." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "AutoTokenizer loading failure and a bare except swallowing errors in FuyuBatchFeature are different defects." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Trainer GPU broadcast failing on NoneType." + }, + { + "left": "issue:23647", + "right": "issue:28218", + "accept": false, + "reason": "Wandb sweeps custom objective and tokenizer added-space behavior are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "Processor/tokenizer access regression is different from Gemma eos_token_id configuration mismatch." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and Whisper empty transcription after align_special_tokens are distinct bugs." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "Hidden_states regression and attention-kernel divergence on torch.compile are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and gather_for_metrics label truncation are different training/generation paths." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "PatchTSMixer post_init policy and non-persistent buffer corruption are not the same issue." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length question on DeBERTa-V3 and missing transformers.onnx module are unrelated." + }, + { + "left": "issue:44521", + "right": "issue:47217", + "accept": false, + "reason": "Both touch multimodal chat/template handling, but all-zero assistant masks and image-placeholder-in-user-text failures are different underlying 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": 28, + "review_unit_index": 40, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:23647", + "issue:25251", + "issue:27957", + "issue:29925", + "issue:35141", + "issue:42490", + "issue:42907", + "issue:43065", + "issue:43352", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43504", + "issue:43525", + "issue:43531", + "issue:43540", + "issue:43646", + "issue:43931", + "issue:44060", + "issue:44162", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:44977", + "issue:45381", + "issue:45412", + "issue:45542", + "issue:45600", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46234", + "issue:46291", + "issue:46365", + "issue:46369", + "issue:46421", + "issue:46640", + "issue:46693", + "issue:47217" + ], + "soft_pairs": [ + "issue:46234|issue:46291", + "issue:43452|issue:44496", + "issue:25251|issue:27957", + "issue:43646|issue:44913", + "issue:44871|issue:46050", + "issue:43352|issue:46640", + "issue:45412|issue:46369", + "issue:43404|issue:43441", + "issue:44464|issue:46421", + "issue:43065|issue:44617", + "issue:43540|issue:44514", + "issue:44162|issue:44792", + "issue:42907|issue:45698", + "issue:46178|issue:46291", + "issue:44792|issue:44977", + "issue:44479|issue:47217", + "issue:44610|issue:46369", + "issue:45381|issue:47217", + "issue:44534|issue:44913", + "issue:45542|issue:46234", + "issue:43525|issue:43531", + "issue:45803|issue:46365", + "issue:44792|issue:45600", + "issue:43504|issue:45698", + "issue:27957|issue:29925", + "issue:46693|issue:47217", + "issue:35141|issue:42490", + "issue:23647|issue:27957", + "issue:45803|issue:46170", + "issue:43931|issue:44060" + ], + "prepared_review_unit_hash": "b6b626aabf7c11ac7c414b1819cf865edfc7186a2bdb2624d45a35906beaf013", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36531, + "estimated_input_tokens": 9133, + "estimated_eval_tokens": 18522 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is thematically broad and noisy: most items are separate Transformers issues around model loading, config serialization, multimodal inputs, or Flash Attention behavior. None of the suggested soft pairs look like true duplicates of the same underlying bug/change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45698 is the best representative: it is concrete, well-described, and has strong discussion activity around a save/load custom-module failure, which fits the main loading/serialization theme in this cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45698 is the most suitable single issue to anchor the cluster because it is specific, reproducible, and broadly representative of the save/load regressions appearing here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Different failures: missing utility symbol vs pipeline import dependency check." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Both are loading errors, but one is gguf_file handling and the other is an unrecognized model config; not the same bug." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "Unrelated: pipeline top_k output shape vs XLM-RoBERTa Flash Attention 2 support." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "Different regressions: custom model init under v5 vs GPTNeoX config reload losing rotary_pct." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Both involve model settings after load, but one is eos_token_id mismatch and the other is quantized sinq save/load breakage." + }, + { + "left": "issue:43352", + "right": "issue:46640", + "accept": false, + "reason": "Specific unsupported model error vs generic Flash Attention 2 issue; not the same code-path." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "One is GPU memory not released after deletion, the other is runtime OOM during Sam3-lite text execution." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Different Mistral/Ministral problems: tied lm_head weights vs FlashAttention failure." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Both touch generation/cache behavior, but one is incorrect outputs and the other is spurious recompilation from lazy cache init." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "Same model family, but dummy Conv2d vs OOM are different defects." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Different multimodal bugs: video processing in Qwen3OmniMoe vs batched chat-template padding crash in Qwen2.5-VL." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage is unrelated to a janus image-generation test failure." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Both are from_pretrained/save_pretrained-adjacent, but one is a trailing-slash path warning and the other loads the wrong custom module." + }, + { + "left": "issue:46178", + "right": "issue:46291", + "accept": false, + "reason": "Different import failures: tensorflow_text backend mapping vs kernels package version/revision requirement." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus image-generation test failure is unrelated to Qwen3.5 Flash Attention generation behavior." + }, + { + "left": "issue:44479", + "right": "issue:47217", + "accept": false, + "reason": "Both are VL regressions, but one is video-input handling after v5.3.0 and the other is image-placeholder placement in user text." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "Different resource/input bugs: processor output size mismatch vs memory overflow in Sam3-lite text." + }, + { + "left": "issue:45381", + "right": "issue:47217", + "accept": false, + "reason": "Both involve VL inputs, but one is wrong vision_position_ids for video input and the other is image placeholder handling." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk in v5 is unrelated to GPTNeoX rotary_pct reload behavior." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "TensorBoard/TensorFlow backend issue is unrelated to a missing falcon_mamba utility function." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Different models and failures: missing pad_token_id vs sliding_window issue." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "Bare except swallowing errors in Fuyu conversion is unrelated to Trainer import conflicts with datasets." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "A failed janus test is not the same as stale auto_mappings references to removed Sam3LiteText configs." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Different loading bugs: legacy field preset loading vs wrong custom-module restoration after save_pretrained." + }, + { + "left": "issue:27957", + "right": "issue:29925", + "accept": false, + "reason": "Flash Attention 2 support for XLM-RoBERTa is unrelated to SigLIP tokenizer fast-mode enforcement." + }, + { + "left": "issue:46693", + "right": "issue:47217", + "accept": false, + "reason": "Different VL/performance problems: FA2 long-context regression vs image-placeholder failure." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "Embedding reinit on resize is unrelated to trailing-slash local path import warnings." + }, + { + "left": "issue:23647", + "right": "issue:27957", + "accept": false, + "reason": "W&B sweeps custom objective is unrelated to XLM-RoBERTa Flash Attention 2." + }, + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Different platforms and bugs: Fuyu exception handling vs Windows ROCm import dependency problems." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Both are Qwen3-family loading/warning issues, but one is a weight-shape mismatch and the other is a tied-weights warning bug; not the same defect." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 41, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:23647", + "issue:25251", + "issue:29127", + "issue:29870", + "issue:35141", + "issue:41950", + "issue:43335", + "issue:43388", + "issue:43475", + "issue:43644", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:43827", + "issue:43854", + "issue:43994", + "issue:44190", + "issue:44242", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44869", + "issue:44913", + "issue:44964", + "issue:45072", + "issue:45127", + "issue:45200", + "issue:45245", + "issue:45381", + "issue:45405", + "issue:45406", + "issue:45584", + "issue:45588", + "issue:45632", + "issue:46082", + "issue:46202", + "issue:46369", + "issue:46399", + "issue:46424", + "issue:46697", + "issue:46832", + "issue:46858", + "issue:46894", + "issue:46932", + "issue:47013", + "issue:47312" + ], + "soft_pairs": [ + "issue:45632|issue:46424", + "issue:45127|issue:45200", + "issue:25251|issue:29870", + "issue:45588|issue:47013", + "issue:44855|issue:45588", + "issue:43673|issue:44743", + "issue:43650|issue:46202", + "issue:41950|issue:43475", + "issue:44242|issue:46894", + "issue:46082|issue:46697", + "issue:44792|issue:45405", + "issue:44617|issue:45584", + "issue:43827|issue:43994", + "issue:43388|issue:43688", + "issue:44190|issue:44964", + "issue:45072|issue:46932", + "issue:29870|issue:35141", + "issue:43644|issue:44913", + "issue:44734|issue:45406", + "issue:44869|issue:46832", + "issue:45245|issue:46858", + "issue:23647|issue:29127", + "issue:43335|issue:44855", + "issue:43854|issue:44792", + "issue:46399|issue:47312", + "issue:45381|issue:46369" + ], + "prepared_review_unit_hash": "b7b2923cfd5707c7eca187a0a7053d6523c416e5df98ca19fb5ca3fb7c166b42", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 26, + "serialized_chars": 35833, + "estimated_input_tokens": 8959, + "estimated_eval_tokens": 18174 + }, + "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 across different models, pipelines, and runtime paths. None of the soft pairs look like true duplicates or mergeable as one fix; at best they share broad keywords like cache, dtype, or serving.", + "confidence": 0.84, + "canonical_issue_reason": "If a representative issue must be chosen, issue:43644 is the cleanest broad regression report: it has a concrete v5.0 behavior change, a core code-path impact, and a clear problem statement. That said, the cluster does not really have a single canonical duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the strongest standalone issue in the set because it describes a specific, high-impact regression with a general scope and a likely actionable fix path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache, but one is a local trust_remote_code path collision and the other is cache max length handling for quantized cache; different bugs and code paths." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "Extended-vocab LoRA collapse vs mm_token_type_ids defaults for Gemma 4 text-only fine-tuning are unrelated model/config bugs." + }, + { + "left": "issue:25251", + "right": "issue:29870", + "accept": false, + "reason": "Top-k pipeline nesting change and BART generate decoder_attention_mask initialization failure are different subsystems and failure modes." + }, + { + "left": "issue:45588", + "right": "issue:47013", + "accept": false, + "reason": "Flash-attention s_aux=None crash and xLSTM return_last_states unpacking bug affect different layers and code paths; not the same fix." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 IndentationError in DebertaV2Model import is unrelated to a flash-attention AttributeError." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Both involve cache, but one is missing GenerationMixin cache in chunked prefill while the other is recurrent-state reset in modular_qwen3_5; distinct defects." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "A placeholder issue and a Gemma pipeline roundtrip bug are clearly unrelated." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Video-classification processor selection and SAM3 vision encoder missing attribute are different vision-stack bugs." + }, + { + "left": "issue:44242", + "right": "issue:46894", + "accept": false, + "reason": "Load-balancing loss gating and ASR training-loss double-shift are both training-loss issues, but they are separate implementation errors." + }, + { + "left": "issue:46082", + "right": "issue:46697", + "accept": false, + "reason": "Llama head_dim validation and Swin2SR over-padding are unrelated config vs preprocessing issues." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A failing janus generate-images test and an unreleased MIN_PEFT_VERSION bump do not share the same underlying bug." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "SAM3 video OOM and Whisper empty-transcription generation failure are unrelated model/runtime problems." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing pipeline() and SIGLIP2 producing bad results with AutoModel/pipeline are different issues; one is documentation, the other runtime behavior." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Metric-gathering truncation and auxiliary-loss normalization are both training-related, but the failure mechanisms and affected code are different." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading and Phi-4 multimodal model loading are not the same bug." + }, + { + "left": "issue:45072", + "right": "issue:46932", + "accept": false, + "reason": "Both mention dtype mismatches, but they affect different models and different quantization/inference paths; too broad to merge." + }, + { + "left": "issue:29870", + "right": "issue:35141", + "accept": false, + "reason": "BART generate uses an uninitialized decoder_attention_mask; resizing embeddings reinitializes output embeddings in post_init. Different lifecycle bugs." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffers filled with junk and GPTNeoXConfig rotary_pct reload defaulting are unrelated state/config issues." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation tensor indexing in serve and Gemma4Processor missing _tokenizer are separate serve-time failures." + }, + { + "left": "issue:44869", + "right": "issue:46832", + "accept": false, + "reason": "Whisper timestamp decoding on trailing replacement chars and RT-DETR forward feature-level mismatch are unrelated model-specific bugs." + }, + { + "left": "issue:45245", + "right": "issue:46858", + "accept": false, + "reason": "Category-count overflow and static-cache seq_length device-tensor arithmetic are different low-level errors." + }, + { + "left": "issue:23647", + "right": "issue:29127", + "accept": false, + "reason": "Wandb sweeps custom objective and LayoutLMv3 error-message clarity do not describe the same issue." + }, + { + "left": "issue:43335", + "right": "issue:44855", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and a Python 3.13 import IndentationError are entirely different problems." + }, + { + "left": "issue:43854", + "right": "issue:44792", + "accept": false, + "reason": "GLM-4.7-Flash loading in tests and janus image-generation test failure are unrelated model/test cases." + }, + { + "left": "issue:46399", + "right": "issue:47312", + "accept": false, + "reason": "Missing decorator in Qwen3.5 rotary registration and kernel version incompatibility are both kernel-related, but not the same concrete bug." + }, + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "Qwen2.5-VL vision_position_ids bug and SAM3 litetext memory overflow are separate model 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": 28, + "review_unit_index": 42, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:39692", + "issue:41720", + "issue:43065", + "issue:43295", + "issue:43317", + "issue:43335", + "issue:43526", + "issue:43540", + "issue:43550", + "issue:43643", + "issue:43673", + "issue:43827", + "issue:43854", + "issue:43994", + "issue:44190", + "issue:44242", + "issue:44479", + "issue:44496", + "issue:44521", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44779", + "issue:44933", + "issue:44964", + "issue:45127", + "issue:45245", + "issue:45276", + "issue:45357", + "issue:45362", + "issue:45375", + "issue:45381", + "issue:45520", + "issue:45563", + "issue:45584", + "issue:45632", + "issue:46170", + "issue:46369", + "issue:46421", + "issue:46489", + "issue:46617", + "issue:46655", + "issue:46697", + "issue:46897", + "issue:47145", + "issue:47312", + "issue:47321" + ], + "soft_pairs": [ + "issue:39692|issue:47321", + "issue:43550|issue:44190", + "issue:45563|issue:45584", + "issue:44617|issue:46369", + "issue:43827|issue:47321", + "issue:43065|issue:44610", + "issue:44479|issue:44734", + "issue:45381|issue:46655", + "issue:41720|issue:45245", + "issue:43526|issue:46697", + "issue:43643|issue:45375", + "issue:39692|issue:44190", + "issue:43335|issue:44190", + "issue:43295|issue:45276", + "issue:44779|issue:46489", + "issue:45632|issue:46617", + "issue:44933|issue:45362", + "issue:43854|issue:44964", + "issue:43540|issue:44734", + "issue:45276|issue:45584", + "issue:43994|issue:47321", + "issue:45127|issue:45357", + "issue:44242|issue:46897", + "issue:47145|issue:47321", + "issue:43673|issue:46421", + "issue:43643|issue:44496", + "issue:45520|issue:46170", + "issue:43317|issue:47312", + "issue:44521|issue:45381" + ], + "prepared_review_unit_hash": "1499fdf516506527dea65384a14507b88cee1f1c35002f460100491d65d7ae2e", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35823, + "estimated_input_tokens": 8956, + "estimated_eval_tokens": 18168 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is largely heterogeneous: most items are unrelated regressions across tokenizers, multimodal processors, config loading, caching, and serving. I found one plausible duplicate pair in the DeepSeek tokenizer regression area; the rest are not the same underlying bug.", + "confidence": 0.82, + "canonical_issue_reason": "No single issue is clearly canonical; the items span multiple unrelated code paths, so there is no strong duplicate anchor for the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a reliable global representative because the cluster lacks one shared root cause.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39692", + "right": "issue:47321", + "accept": false, + "reason": "SigLIP2 docs/model mismatch and AMD quark class updates are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "A torch.compile/SDPA model bug is unrelated to a local dataset loading script failure." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "Paged generate warning handling and Whisper empty-transcription failure are different generation bugs." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Both are memory-related in Sam3, but they describe different model variants and failure modes, not the same bug." + }, + { + "left": "issue:43827", + "right": "issue:47321", + "accept": false, + "reason": "Docs referencing pipeline() after v5 removals is unrelated to AMD quark class updates." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "Sam3PixelDecoder's dummy Conv2d and OmDet-Turbo image-size mismatch are different model/processor issues." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "A video-input regression in Qwen VL models is unrelated to a serving KV-cache tensor indexing crash." + }, + { + "left": "issue:45381", + "right": "issue:46655", + "accept": false, + "reason": "Wrong vision_position_ids in qwen2.5-vl and SmolVLM video padding are different multimodal processor bugs." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA asserts and the 2^24 category limit are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:46697", + "accept": false, + "reason": "BEiT label reduction and Swin2SR over-padding affect different processors and code paths." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "Generic AutoConfig field loss under trust_remote_code and Qwen3.5MoeVisionConfig dropping one field are similar in theme but not the same concrete bug." + }, + { + "left": "issue:39692", + "right": "issue:44190", + "accept": false, + "reason": "SigLIP2 example errors are unrelated to loading a local dataset in the image-classification script." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation and dataset loading in a training script are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "Processor.tokenizer regression and resize_token_embeddings behavior are different API issues." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": true, + "reason": "Both report the same DeepSeek tokenizer regression in v5+ producing incorrect outputs; 46489 reads like a specific follow-up to the broader 44779 report." + }, + { + "left": "issue:45632", + "right": "issue:46617", + "accept": false, + "reason": "Both involve trust_remote_code caching, but one is a leaf-directory collision and the other is a cache FileNotFoundError; not enough to treat as the same bug." + }, + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "A missing import from image_utils is unrelated to a Qwen3.5 chat crash." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "These are separate model-loading failures for different models with no shared root cause shown." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Qwen3OmniMoe video-input validation and serving KV-cache indexing are different code paths." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "Embedding resize propagation and Whisper empty-transcription generation are unrelated bugs." + }, + { + "left": "issue:43994", + "right": "issue:47321", + "accept": false, + "reason": "SigLIP2 nonsensical outputs and AMD quark class updates are unrelated." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "LoRA merge with extended vocabulary and saving Qwen3.5 visual encoder keys are different model-management issues." + }, + { + "left": "issue:44242", + "right": "issue:46897", + "accept": false, + "reason": "MoE load-balancing loss and Florence2 double-shift training loss are unrelated losses." + }, + { + "left": "issue:47145", + "right": "issue:47321", + "accept": false, + "reason": "FP8 weight_scale loading and AMD quark class updates are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:46421", + "accept": false, + "reason": "Chunked-prefill cache missing vs. chunked-prefill recompilation are related in theme but not clearly the same concrete defect." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "Missing fields from trust_remote_code config loading and an unrecognized model_type error are different config-loading failures." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Python 3.13 flash_attn import handling and Windows ROCm torch.distributed dependency issues are unrelated." + }, + { + "left": "issue:43317", + "right": "issue:47312", + "accept": false, + "reason": "Dequantized model offload loading and kernel version compatibility are separate problems." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "All-zero assistant masks and wrong vision_position_ids are different multimodal template/positioning bugs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet and stays appropriately conservative. The cluster does appear heterogeneous from the titles alone, and the one accepted duplicate pair (44779/46489) is plausibly the same DeepSeek tokenizer regression. The other soft-edge rejections are cautious and well-supported by the visible issue descriptions." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 43, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:29925", + "issue:32090", + "issue:35141", + "issue:36246", + "issue:38617", + "issue:41628", + "issue:42175", + "issue:42796", + "issue:43295", + "issue:43335", + "issue:43404", + "issue:43531", + "issue:43611", + "issue:43673", + "issue:43976", + "issue:44016", + "issue:44112", + "issue:44246", + "issue:44279", + "issue:44462", + "issue:44496", + "issue:44521", + "issue:44737", + "issue:44938", + "issue:44964", + "issue:45003", + "issue:45042", + "issue:45083", + "issue:45216", + "issue:45290", + "issue:45335", + "issue:45542", + "issue:45593", + "issue:45600", + "issue:45632", + "issue:45759", + "issue:45803", + "issue:45907", + "issue:46234", + "issue:46506", + "issue:46752", + "issue:46858", + "issue:47217", + "issue:47321", + "issue:47448" + ], + "soft_pairs": [ + "issue:44279|issue:47448", + "issue:43611|issue:47321", + "issue:44112|issue:45600", + "issue:44016|issue:47448", + "issue:42175|issue:44279", + "issue:44112|issue:44964", + "issue:43673|issue:46858", + "issue:44246|issue:45003", + "issue:44462|issue:45907", + "issue:27957|issue:35141", + "issue:42796|issue:44521", + "issue:32090|issue:35141", + "issue:44964|issue:45600", + "issue:44737|issue:45803", + "issue:43404|issue:45216", + "issue:43531|issue:45600", + "issue:43335|issue:45593", + "issue:25251|issue:29925", + "issue:44496|issue:47217", + "issue:45290|issue:46752", + "issue:44496|issue:45042", + "issue:45632|issue:46234", + "issue:43976|issue:47448", + "issue:36246|issue:41628", + "issue:36246|issue:38617", + "issue:43295|issue:45335", + "issue:45083|issue:46506", + "issue:44496|issue:45759", + "issue:44938|issue:45542" + ], + "prepared_review_unit_hash": "811de344b87636eddef3ea5c32840cffebbedac44539673a336b9b432758c456", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35899, + "estimated_input_tokens": 8975, + "estimated_eval_tokens": 18206 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: the issues span unrelated model-loading, tokenizer, chat-template, packaging, CI, and training bugs. None of the soft pairs look like the same underlying defect, so no duplicates should be merged here.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue; the cluster is not a true duplicate set and mixes many unrelated failure modes.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue cleanly represents the cluster, since the items cover different subsystems and distinct root causes.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44279", + "right": "issue:47448", + "accept": false, + "reason": "Generic dependency/install complaint vs a specific Gemma4 architecture recognition bug." + }, + { + "left": "issue:43611", + "right": "issue:47321", + "accept": false, + "reason": "Base-model-prefix loading regression is unrelated to AMD quark class refresh." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "CI failure in GraniteSpeech device override is unrelated to stale auto_mappings entries." + }, + { + "left": "issue:44016", + "right": "issue:47448", + "accept": false, + "reason": "Notebook syntax error is documentation-related, not a Gemma4 model-loading issue." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "TensorFlow missing from the torch extra is a concrete packaging bug, not a vague dependency issue." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "GraniteSpeech CI flake and Phi-4 multimodal loading failure are different problems." + }, + { + "left": "issue:43673", + "right": "issue:46858", + "accept": false, + "reason": "Missing GenerationMixin cache in v5.0.0 is distinct from device-tensor slice arithmetic in prefill." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Slow import reports a performance issue; unsafe sys.modules access is a correctness bug." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer.json being ignored is unrelated to forwarding revision in list_repo_files." + }, + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 support for XLM-R is unrelated to resize-token-embeddings reinitializing output weights." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "BPE marker handling in a tokenizer is different from assistant_mask generation for multimodal chat templates." + }, + { + "left": "issue:32090", + "right": "issue:35141", + "accept": false, + "reason": "Trainer broadcast NoneType error is unrelated to post-init embedding reinitialization." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Phi-4 multimodal load failure and stale auto_mappings references are different load paths and causes." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "CPU arange in XLNet is a performance/device bug, not a bare-except exception-handling issue." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Untied lm_head weights in Mistral3 is distinct from Qwen3.5 checkpoint save corruption." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior is unrelated to removed Sam3LiteText config references." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "Sparse-layer creation in SwitchTransformers is unrelated to D-FINE auxiliary loss behavior." + }, + { + "left": "issue:25251", + "right": "issue:29925", + "accept": false, + "reason": "Pipeline top_k nesting bug is unrelated to SigLIP fast-tokenizer enforcement." + }, + { + "left": "issue:44496", + "right": "issue:47217", + "accept": false, + "reason": "Unrecognized model type error is unrelated to VL image-placeholder handling in user text." + }, + { + "left": "issue:45290", + "right": "issue:46752", + "accept": false, + "reason": "Both hit apply_chat_template, but one is tool-call/no-content handling and the other is empty-input IndexError." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Model-type resolution failure is unrelated to PIL image processors wrongly requiring torchvision." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "Local cache path collisions are unrelated to a missing import_utils availability helper." + }, + { + "left": "issue:43976", + "right": "issue:47448", + "accept": false, + "reason": "Python version compatibility breakage is unrelated to Gemma4 architecture registration." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Missing Qwen2_5_VLImageProcessor export is a different import failure than AutoImageProcessor absence." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "A missing processor export and a missing configuration_utils symbol are unrelated import errors." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Processor/tokenizer regression is unrelated to t5gemma resize_token_embeddings not affecting decoder embeddings." + }, + { + "left": "issue:45083", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3_omni_moe feature-length helper behavior is unrelated to PPDocLayoutV3 CPU/CUDA divergence." + }, + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Model-type recognition failure is not the same as unwrapping text_config for composite multimodal configs." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import failure is unrelated to tensorboard installing without TensorFlow." + } + ] + }, + "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": 28, + "review_unit_index": 44, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29925", + "issue:33290", + "issue:39692", + "issue:41950", + "issue:42796", + "issue:42907", + "issue:42994", + "issue:43295", + "issue:43381", + "issue:43441", + "issue:43526", + "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:44683", + "issue:44779", + "issue:44792", + "issue:44933", + "issue:44987", + "issue:45020", + "issue:45335", + "issue:45405", + "issue:45446", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45812", + "issue:45823", + "issue:46082", + "issue:46121", + "issue:46424", + "issue:46506", + "issue:46515", + "issue:46752", + "issue:46932", + "issue:47217", + "issue:47332", + "issue:47333" + ], + "soft_pairs": [ + "issue:39692|issue:46932", + "issue:41950|issue:44016", + "issue:43976|issue:44987", + "issue:44279|issue:46515", + "issue:44683|issue:45446", + "issue:42994|issue:47333", + "issue:29925|issue:33290", + "issue:43673|issue:46424", + "issue:43673|issue:44336", + "issue:43381|issue:45593", + "issue:43526|issue:43901", + "issue:44295|issue:46506", + "issue:44016|issue:44220", + "issue:46082|issue:46121", + "issue:44327|issue:46752", + "issue:43650|issue:44220", + "issue:42994|issue:47332", + "issue:44016|issue:44062", + "issue:43441|issue:45588", + "issue:44792|issue:46424", + "issue:41950|issue:45020", + "issue:42907|issue:43441", + "issue:45335|issue:45584", + "issue:43824|issue:44933", + "issue:44279|issue:45405", + "issue:39692|issue:43901", + "issue:43901|issue:44496", + "issue:43295|issue:44779", + "issue:44779|issue:45812", + "issue:42796|issue:43650", + "issue:43901|issue:47217", + "issue:41950|issue:45823" + ], + "prepared_review_unit_hash": "1220ee94181e8b725af5ad9d3b9b038c67dbfd652cdedc7aacc4fab21814fd15", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 36916, + "estimated_input_tokens": 9229, + "estimated_eval_tokens": 18714 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the titles point to many unrelated bugs, regressions, and doc issues across different models/subsystems. I rejected all soft edges; none look like the same underlying change or bug.", + "confidence": 0.94, + "canonical_issue_reason": "#45020 is the broadest, most general regression report in the set and is better suited as a representative issue than the narrower model- or feature-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "#45020 has the strongest cluster-representative value: it is a clearly described, high-signal regression with broader scope than the other issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39692", + "right": "issue:46932", + "accept": false, + "reason": "Different models and different failures: SigLIP2 doc example errors vs Gemma4 quantized embedder dtype casting." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Unrelated bugs: a video-classification processor lookup issue vs a notebook syntax error." + }, + { + "left": "issue:43976", + "right": "issue:44987", + "accept": false, + "reason": "Both involve version regressions, but one is a Python support issue and the other is a model-loading failure for a specific repo; not the same bug." + }, + { + "left": "issue:44279", + "right": "issue:46515", + "accept": false, + "reason": "A dependency issue with transformers is not the same as a torch version incompatibility." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Same subsystem (flex_attention) but different concrete problems: compiled flex_attention on torch>=2.9 vs a bad version check for AuxRequest import." + }, + { + "left": "issue:42994", + "right": "issue:47333", + "accept": false, + "reason": "Both are save failures, but they involve different model families and different save contexts (dequantized vs disk-offloaded)." + }, + { + "left": "issue:29925", + "right": "issue:33290", + "accept": false, + "reason": "Tokenizer fast-mode enforcement is unrelated to an Adafactor/DeepSpeed OOM." + }, + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache, but one is a missing GenerationMixin cache in chunked prefill and the other is cache-length behavior for quantized cache only." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Cache-missing regression and ANSI output formatting are unrelated." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing in eval mode is unrelated to D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "Different components and bug types: BeitImageProcessorFast label reduction vs documentation drift in TextClassificationPipeline." + }, + { + "left": "issue:44295", + "right": "issue:46506", + "accept": false, + "reason": "A position_ids buffer access error is unrelated to CPU/CUDA output differences in PPDocLayoutV3." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "A notebook syntax error is unrelated to the fbank feature extraction bug." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "Both are config-validation issues, but they concern different fields and different code paths." + }, + { + "left": "issue:44327", + "right": "issue:46752", + "accept": false, + "reason": "Both are edge-case crashes, but in different APIs and with different failure modes." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "The placeholder issue 'ADD THE DATA' is not the same as the fbank feature extraction failure." + }, + { + "left": "issue:42994", + "right": "issue:47332", + "accept": false, + "reason": "Same general saving area, but distinct models and distinct offloading/save scenarios." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken keyword collision are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both touch flash attention, but one is a model-specific v5 RC failure and the other is an integration AttributeError on sink-less models." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "A janus test failure is unrelated to cache-length semantics." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Video-classification processor lookup and remote_code loading regressions are different bugs." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Saving dequantized Ministrals/Devstrals is unrelated to Ministral-3 FlashAttention failures." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "Tokenizer embedding resize behavior and Whisper empty transcription generation are different issues." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they concern different missing symbols in different modules." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "A transformers dependency issue is unrelated to an unreleased PEFT minimum version bump." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "A SigLIP2 docs/model mismatch is unrelated to a pipeline documentation note about return_all_scores." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Documentation drift in a pipeline API is unrelated to an unrecognized-model config error." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Both are regression reports around tokenizers, but they describe different models and different breakages." + }, + { + "left": "issue:43901", + "right": "issue:47217", + "accept": false, + "reason": "A pipeline docs issue is unrelated to VL models failing on image placeholders in user text." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Video-classification processor lookup and Gemma4 device-map mismatch are different code paths and failures." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 45, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:43299", + "issue:43381", + "issue:43479", + "issue:43577", + "issue:43644", + "issue:43650", + "issue:43653", + "issue:43854", + "issue:43873", + "issue:43931", + "issue:43976", + "issue:44016", + "issue:44075", + "issue:44242", + "issue:44291", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44466", + "issue:44683", + "issue:45081", + "issue:45103", + "issue:45200", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45357", + "issue:45372", + "issue:45375", + "issue:45405", + "issue:45542", + "issue:45588", + "issue:45593", + "issue:45800", + "issue:45859", + "issue:45907", + "issue:46082", + "issue:46123", + "issue:46209", + "issue:46369", + "issue:46399", + "issue:46512", + "issue:46614", + "issue:46851", + "issue:47000", + "issue:47085", + "issue:47145", + "issue:47405" + ], + "soft_pairs": [ + "issue:43299|issue:45216", + "issue:43479|issue:45375", + "issue:44466|issue:45276", + "issue:46123|issue:46369", + "issue:46614|issue:47085", + "issue:43577|issue:45216", + "issue:43653|issue:45593", + "issue:47145|issue:47405", + "issue:43976|issue:45405", + "issue:43381|issue:47000", + "issue:44462|issue:45081", + "issue:46082|issue:46399", + "issue:43299|issue:45357", + "issue:43299|issue:43854", + "issue:43644|issue:47145", + "issue:44075|issue:44368", + "issue:46082|issue:46512", + "issue:44242|issue:45200", + "issue:44016|issue:44291", + "issue:44683|issue:45542", + "issue:43931|issue:47145", + "issue:43650|issue:46209", + "issue:45588|issue:47000", + "issue:45081|issue:45907", + "issue:45859|issue:46851", + "issue:45103|issue:45372", + "issue:43873|issue:47145", + "issue:44351|issue:45278", + "issue:45542|issue:45800" + ], + "prepared_review_unit_hash": "45b4e926bacd89f73536fe4757321e39d1fbe4e5ab61f3a3cd4c232b9cde45de", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36773, + "estimated_input_tokens": 9194, + "estimated_eval_tokens": 18644 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: most items are distinct model/config/load/save/tokenizer bugs rather than duplicates. No soft pair looks like the same underlying change or failure mode, so none should be merged together.", + "confidence": 0.93, + "canonical_issue_reason": "No true canonical duplicate stands out; if one must be chosen, issue 43299 is the broadest early regression report and the closest thing to a hub, but it is still distinct from the others.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43299 is the most central candidate only in the sense that it is an early, user-facing regression report with several superficial similarity links; however it does not actually subsume the other issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Both mention Qwen regressions, but one is a loading failure and the other is incorrect save_pretrained serialization; different code paths and symptoms." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Different models and different config bugs: Phi4Multimodal default init vs Qwen3_5MoeVisionConfig field dropping." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "One is tied-weight serialization varying by device, the other is Gemma 4 embedding resize propagation; not the same defect." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "MaskGenerationPipeline batch-finalization bug vs sam3-litetext memory overflow; unrelated failures." + }, + { + "left": "issue:46614", + "right": "issue:47085", + "accept": false, + "reason": "Stale phonemizer backend state is unrelated to Rust tokenizer thread safety." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Blip2 dtype propagation bug and Qwen3.5 checkpoint save regression are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Tokenizer special-token registration bug vs D-FINE auxiliary-loss behavior; no shared underlying bug." + }, + { + "left": "issue:47145", + "right": "issue:47405", + "accept": false, + "reason": "Both are silent load/corruption issues, but one is FP8 weight_scale application and the other is wrong Auto class/model_type conversion skip; different mechanics." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version support issue vs unreleased MIN_PEFT_VERSION bump; unrelated." + }, + { + "left": "issue:43381", + "right": "issue:47000", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction is unrelated to RTDetrLoss infeasible cost matrix errors." + }, + { + "left": "issue:44462", + "right": "issue:45081", + "accept": false, + "reason": "AutoTokenizer ignoring tokenizer.json is a separate loading bug from Mistral regex patch crashing on missing backend_tokenizer." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "LlamaConfig head_dim validation and Qwen3.5 rotary-kernel registration failure are different issues." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5/Qwen3VL regressions, but one is model loading and the other is save_pretrained visual key serialization; not the same bug." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Different Qwen variants and different failure modes: model loading mismatch vs unit-test load failure." + }, + { + "left": "issue:43644", + "right": "issue:47145", + "accept": false, + "reason": "Non-persistent buffer junk in Transformers 5.0 is unrelated to FP8 compressed-tensors weight_scale loading." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Optimizer SGD args ignored and tie_word_embeddings warning are unrelated; no shared code-path." + }, + { + "left": "issue:46082", + "right": "issue:46512", + "accept": false, + "reason": "Both involve model config parameters, but one is Llama head_dim validation and the other is from_config dtype staleness." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "Load-balancing loss gating and Gemma 4 mm_token_type_ids defaulting are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error and init_empty_weights unexpected argument are unrelated." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Torch 2.9 flex_attention compilation failure is unrelated to tensorboard/TensorFlow backend installation behavior." + }, + { + "left": "issue:43931", + "right": "issue:47145", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch and FP8 weight_scale omission are different loading failures." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Placeholder/non-technical issue vs FP8 inference failure on B200; no relation." + }, + { + "left": "issue:45588", + "right": "issue:47000", + "accept": false, + "reason": "Flash-attention auxiliary state crash and RTDetrLoss infeasible cost matrix are different subsystems and bugs." + }, + { + "left": "issue:45081", + "right": "issue:45907", + "accept": false, + "reason": "Mistral regex tokenizer crash vs list_repo_files revision forwarding; unrelated." + }, + { + "left": "issue:45859", + "right": "issue:46851", + "accept": false, + "reason": "Both are Qwen3.5 rotary/position-related, but one is CPU offload compatibility and the other is variable-length position_ids handling; not the same concrete bug." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "auto_docstring annotation handling and mistral_common import failure are unrelated." + }, + { + "left": "issue:43873", + "right": "issue:47145", + "accept": false, + "reason": "Quantization/offloading behavior and FP8 frozen checkpoint weight_scale application are distinct loading paths." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "A single missing HybridCache import is not the same as broad import errors after upgrade." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "TensorBoard without TensorFlow backend error and torch/transformers incompatibility 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": 28, + "review_unit_index": 46, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:33290", + "issue:42175", + "issue:42796", + "issue:42994", + "issue:43317", + "issue:43441", + "issue:43479", + "issue:43525", + "issue:43575", + "issue:43643", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44162", + "issue:44186", + "issue:44188", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44336", + "issue:44373", + "issue:44496", + "issue:44589", + "issue:44623", + "issue:44671", + "issue:44749", + "issue:44779", + "issue:45325", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45588", + "issue:45600", + "issue:45656", + "issue:45712", + "issue:45834", + "issue:46399", + "issue:46846", + "issue:46851", + "issue:47000", + "issue:47145", + "issue:47217", + "issue:47333", + "issue:47427" + ], + "soft_pairs": [ + "issue:43441|issue:45834", + "issue:42994|issue:43317", + "issue:44336|issue:46851", + "issue:44671|issue:44779", + "issue:43784|issue:45542", + "issue:44186|issue:47000", + "issue:44246|issue:47217", + "issue:42175|issue:43976", + "issue:45325|issue:46851", + "issue:43673|issue:43761", + "issue:44623|issue:44749", + "issue:46399|issue:46846", + "issue:43525|issue:43643", + "issue:43479|issue:43525", + "issue:44589|issue:45362", + "issue:43901|issue:44246", + "issue:44623|issue:45397", + "issue:42796|issue:43441", + "issue:33290|issue:45656", + "issue:43575|issue:47145", + "issue:44261|issue:44263", + "issue:44188|issue:44336", + "issue:45600|issue:45712", + "issue:45588|issue:47427", + "issue:45362|issue:47333", + "issue:43824|issue:43976", + "issue:44246|issue:44496", + "issue:44336|issue:44373", + "issue:43723|issue:44162" + ], + "prepared_review_unit_hash": "8eaf943b6ea439728105dd94115c2e7a5c665a4830249ecbb9887019da0e80a7", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36258, + "estimated_input_tokens": 9065, + "estimated_eval_tokens": 18386 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most items are unrelated model-, tokenizer-, packaging-, or runtime-specific bugs that only share broad Transformers/v5 wording. I do not see a safe duplicate merge candidate among the soft edges.", + "confidence": 0.21, + "canonical_issue_reason": "If a representative issue is needed, issue 44246 is the broadest/highest-signal runtime problem in the set, but the overall cluster is not a coherent duplicate family.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44246 is the most representative general-runtime complaint among otherwise unrelated reports, though it should not be treated as a true cluster duplicate anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "Different bugs: Ministral tokenizer BPE-marker handling vs Kosmos2.5 OCR long-input index error." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Saving a quantized model and loading a dequantized model with offload are different code paths and failures." + }, + { + "left": "issue:44336", + "right": "issue:46851", + "accept": false, + "reason": "ANSI-code leakage in loading_report is unrelated to Qwen3.5 variable-length position_ids handling." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "CamemBERT masked-LM regression and Deepseek tokenizer corruption are distinct model/tokenizer problems." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both are import/backend errors, but they affect different packages and stem from different missing symbols/backends." + }, + { + "left": "issue:44186", + "right": "issue:47000", + "accept": false, + "reason": "Tokenizer crash on NER/padding is not the same underlying bug as RTDetrLoss infeasible cost matrix." + }, + { + "left": "issue:44246", + "right": "issue:47217", + "accept": false, + "reason": "Import-time slowness and VL image-placeholder failure are unrelated behaviors." + }, + { + "left": "issue:42175", + "right": "issue:43976", + "accept": false, + "reason": "Package extra dependency inclusion differs from Python-version compatibility; not the same defect." + }, + { + "left": "issue:45325", + "right": "issue:46851", + "accept": false, + "reason": "Qwen2.5-VL rope-index scaling and Qwen3.5 variable-length position_ids are separate model bugs." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Missing generation cache in chunked_prefill is unrelated to CLIPVision hidden_states regression." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "Missing save_pretrained files and slower filtering after upgrade are different symptoms and code paths." + }, + { + "left": "issue:46399", + "right": "issue:46846", + "accept": false, + "reason": "A kernel-registration crash from a missing decorator is not the same as TP-plan OOM/shape mismatch." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "A missing pad_token_id attribute in Llama4Config is different from trust_remote_code dropping fields in AutoConfig." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Multimodal default-config initialization and missing config attributes are unrelated configuration bugs." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Float8 storage lookup failure and Qwen3.5 chat crash do not share a concrete code path." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "A documentation mismatch is not the same as import performance degradation." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "Processor save output omissions and gemma-4 zero3 from_pretrained failures are distinct issues." + }, + { + "left": "issue:42796", + "right": "issue:43441", + "accept": false, + "reason": "Tokenizer BPE-marker parsing and FlashAttention failure are different parts of the stack." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "AdaFactor OOM in deepspeed and double optimizer-step behavior are not the same underlying bug." + }, + { + "left": "issue:43575", + "right": "issue:47145", + "accept": false, + "reason": "Tensor-parallel load OOM and FP8 checkpoint weight_scale omission are different loading bugs." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "MLA q_a_layernorm epsilon handling is unrelated to torch.split return-value behavior." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "torch.compile-dependent attention-kernel divergence has nothing to do with ANSI output in loading_report." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Stale auto_mappings entries and leftover dummy_pt_objects cleanup are separate maintenance failures." + }, + { + "left": "issue:45588", + "right": "issue:47427", + "accept": false, + "reason": "FlashAttention sink-less-model crash and redundant init on non-CUDA hardware are different initialization paths." + }, + { + "left": "issue:45362", + "right": "issue:47333", + "accept": false, + "reason": "Qwen3.5 chat crashes and disk-offloaded checkpoint saving are unrelated failures." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "An import error for a missing class and a Python support/package issue are not duplicates." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import slowness is unrelated to unrecognized-model/config.json validation failure." + }, + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI-code generation and a wrong docstring are not the same issue." + }, + { + "left": "issue:43673", + "right": "issue:44162", + "accept": false, + "reason": "Generation cache regression and ESM2 breakage are entirely different model/runtime bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 47, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:36246", + "issue:42491", + "issue:42915", + "issue:43257", + "issue:43299", + "issue:43352", + "issue:43526", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:44246", + "issue:44295", + "issue:44297", + "issue:44336", + "issue:44479", + "issue:44779", + "issue:44871", + "issue:45003", + "issue:45072", + "issue:45083", + "issue:45103", + "issue:45137", + "issue:45250", + "issue:45292", + "issue:45375", + "issue:45397", + "issue:45561", + "issue:45563", + "issue:45588", + "issue:45701", + "issue:45759", + "issue:46291", + "issue:46399", + "issue:46424", + "issue:46506", + "issue:46614", + "issue:46655", + "issue:46665", + "issue:46726", + "issue:46752", + "issue:46821", + "issue:46829", + "issue:46851", + "issue:47176", + "issue:47177", + "issue:47217", + "issue:47312" + ], + "soft_pairs": [ + "issue:42491|issue:43299", + "issue:46291|issue:46399", + "issue:46399|issue:46851", + "issue:45072|issue:46829", + "issue:46655|issue:46726", + "issue:45083|issue:46851", + "issue:44297|issue:45397", + "issue:47176|issue:47177", + "issue:45003|issue:47217", + "issue:43526|issue:45003", + "issue:45588|issue:46851", + "issue:45003|issue:46821", + "issue:45103|issue:46399", + "issue:44246|issue:44871", + "issue:43550|issue:46506", + "issue:45561|issue:45563", + "issue:44336|issue:46424", + "issue:44295|issue:45397", + "issue:43540|issue:46752", + "issue:42915|issue:47312", + "issue:36246|issue:43596", + "issue:43540|issue:46655", + "issue:46614|issue:46665", + "issue:44479|issue:46655", + "issue:46726|issue:47217", + "issue:45375|issue:45759", + "issue:43257|issue:45137", + "issue:43352|issue:45250", + "issue:45292|issue:45701", + "issue:44779|issue:45292" + ], + "prepared_review_unit_hash": "41e7f001a0ce52873c9f033f53014f5ea2f666be4fad3788bf2b2168d78dee4f", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36198, + "estimated_input_tokens": 9050, + "estimated_eval_tokens": 18356 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly separate bugs across different model families and subsystems; none of the soft pairs looks like a safe duplicate merge on the same concrete code path. The only strong commonality is broad Qwen/VL regression chatter, but the reported failures are still materially different.", + "confidence": 0.74, + "canonical_issue_reason": "issue:44479 is the broadest concrete regression report here: it names a specific version breakage and multiple affected Qwen VL/MoE variants, making it the best anchor for nearby vision-regression reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best representative issue in this set because it is self-contained, version-specific, and spans multiple related model variants; the others are either narrower, unrelated, or clearly different bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both are Qwen3 MoE loading regressions, but one is a LoRA compatibility problem and the other is a base-model load failure; not the same concrete bug." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "One is a kernels import/version requirement failure, the other is a Qwen3.5 kernel registration error; same area, different failure mode and fix." + }, + { + "left": "issue:46399", + "right": "issue:46851", + "accept": false, + "reason": "Different problems in Qwen3.5: kernels registration vs variable-length position_ids handling; not mergeable as one bug." + }, + { + "left": "issue:45072", + "right": "issue:46829", + "accept": false, + "reason": "Unrelated model/path issues: bf16 dtype mismatches in SwitchTransformers/TimmWrapperModel versus VideoPrism output formatting." + }, + { + "left": "issue:46655", + "right": "issue:46726", + "accept": false, + "reason": "Both involve vision processing, but one is a video padding bug and the other is missing-backend AutoProcessor error handling." + }, + { + "left": "issue:45083", + "right": "issue:46851", + "accept": false, + "reason": "Both touch Qwen input-shape logic, but one is feature-extraction length math and the other is position_ids with variable-length inputs." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to Gemma-4 ZeRO3 from_pretrained behavior." + }, + { + "left": "issue:47176", + "right": "issue:47177", + "accept": false, + "reason": "Both are security issues, but command-substitution allowlist bypass and tar/zip slip arbitrary write are distinct vulnerabilities with different code paths." + }, + { + "left": "issue:45003", + "right": "issue:47217", + "accept": false, + "reason": "sys.modules access in modeling_utils is unrelated to VL placeholder handling in user text." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to unsafe sys.modules access." + }, + { + "left": "issue:45588", + "right": "issue:46851", + "accept": false, + "reason": "Flash-attention sinkless-model crash and Qwen3.5 variable-length position_ids are different defects." + }, + { + "left": "issue:45003", + "right": "issue:46821", + "accept": false, + "reason": "Generic modeling_utils sys.modules access does not match TimesFM window_size AttributeError." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "auto_docstring future-annotations crash is unrelated to Qwen3.5 kernel registration." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import-time performance regression and Gemma eos_token_id inconsistency are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:46506", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba is unrelated to CPU/CUDA divergence in PPDocLayoutV3." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "pytest-xdist file race and stale generate() warning are different test/behavior issues." + }, + { + "left": "issue:44336", + "right": "issue:46424", + "accept": false, + "reason": "ANSI coloring in loading_report and cache max length handling are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "position_ids buffer access and Gemma-4 ZeRO3 from_pretrained are separate problems." + }, + { + "left": "issue:43540", + "right": "issue:46752", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and empty chat list IndexError are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:47312", + "accept": false, + "reason": "FineGrainedFP8Config failure and kernels 0.16 incompatibility are different versioning/runtime issues." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "Qwen2_5_VLImageProcessor import error and BertModel ZeRO3 index error are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:46655", + "accept": false, + "reason": "Both are video-related, but they affect different models and different processing bugs." + }, + { + "left": "issue:46614", + "right": "issue:46665", + "accept": false, + "reason": "Stale phonemizer backend state is unrelated to inconsistent from_pretrained behavior inside Python classes." + }, + { + "left": "issue:44479", + "right": "issue:46655", + "accept": false, + "reason": "Both are video-input regressions, but one is a broad Qwen version regression and the other is a SmolVLM video padding bug." + }, + { + "left": "issue:46726", + "right": "issue:47217", + "accept": false, + "reason": "Missing vision backend handling is unrelated to VL placeholder placement in user text." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Missing deepstack_visual_indexes and text_config unwrapping are separate Qwen multimodal config issues." + }, + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both involve DeepSpeed/ZeRO3, but one is Qwen3 MoE weight conversion and the other is a deque underflow during initialization." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "A model not yet supporting Flash Attention 2.0 is not the same as the generic Flash Attention 2.0 issue." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Token embedding resizing and tokenizer-version drift are unrelated." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Deepseek tokenizer regression and resize_token_embeddings/output_embeddings mismatch are different bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 48, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:36246", + "issue:42757", + "issue:42886", + "issue:43295", + "issue:43475", + "issue:43540", + "issue:44038", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44242", + "issue:44295", + "issue:44351", + "issue:44368", + "issue:44451", + "issue:44466", + "issue:44479", + "issue:44623", + "issue:44749", + "issue:44857", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45459", + "issue:45561", + "issue:45588", + "issue:45593", + "issue:45600", + "issue:45685", + "issue:46132", + "issue:46170", + "issue:46369", + "issue:46431", + "issue:46506", + "issue:46726", + "issue:46752", + "issue:47217", + "issue:47312", + "issue:47321", + "issue:47429", + "issue:47455" + ], + "soft_pairs": [ + "issue:44242|issue:45593", + "issue:43295|issue:45292", + "issue:44995|issue:45200", + "issue:45685|issue:46431", + "issue:44060|issue:44077", + "issue:46170|issue:47321", + "issue:44162|issue:45200", + "issue:44466|issue:45335", + "issue:36246|issue:44351", + "issue:36246|issue:42757", + "issue:45137|issue:45600", + "issue:42886|issue:46132", + "issue:44995|issue:45588", + "issue:44451|issue:46752", + "issue:44479|issue:46170", + "issue:44295|issue:44749", + "issue:44877|issue:45375", + "issue:47217|issue:47429", + "issue:43475|issue:46369", + "issue:44877|issue:46170", + "issue:45003|issue:45561", + "issue:44987|issue:45459", + "issue:44623|issue:46726", + "issue:47312|issue:47455", + "issue:43475|issue:43540", + "issue:44242|issue:44857", + "issue:44038|issue:46506", + "issue:44368|issue:44977" + ], + "prepared_review_unit_hash": "60d5d030ccbcb5ee435a69905eb8b571be9e2346b254003801cb16c629ce264f", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36000, + "estimated_input_tokens": 9000, + "estimated_eval_tokens": 18256 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: most items are unrelated model- or subsystem-specific regressions, not true duplicates. The only plausible duplicate pair is the kernels version-checking reports.", + "confidence": 0.63, + "canonical_issue_reason": "issue:47455 is the cleanest root-cause report: it points to a broken kernel version check, which can explain the compatibility failure raised in issue:47312.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:47455 is the best representative issue because it describes the underlying version-check bug more directly than the symptom report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Different models and loss paths; both mention missing losses, but they are not the same bug or code path." + }, + { + "left": "issue:43295", + "right": "issue:45292", + "accept": false, + "reason": "One is a processor/tokenizer regression, the other is embedding resizing not propagating; different failures." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "Unrelated model bugs: stale cache on second forward vs missing mm_token_type_ids default." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "MPS histogram/int error and missing chat template are unrelated processor/runtime issues." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Both are warning/config complaints, but they involve different models and different checks." + }, + { + "left": "issue:46170", + "right": "issue:47321", + "accept": false, + "reason": "Different AMD-related problems: torch.distributed dependency vs quark class update." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma 4 token-type defaults are unrelated model-specific regressions." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both involve embeddings, but one is serialization of tied weights while the other is resize_token_embeddings propagation; not the same concrete bug." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Different missing import errors for different symbols and packages." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "Separate ImportErrors from different modules; no shared root cause." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "DeepSpeed deque failure and auto_mappings CI breakage are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and AttentionInterface.register behavior change are different issues." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "Stale indexer cache vs flash-attention s_aux=None crash are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:46752", + "accept": false, + "reason": "Model loading failure and empty-list chat template crash are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "Video-input regression in VL models is unrelated to Windows/ROCm distributed import issues." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "Position_ids buffer access error and performance slowdown after upgrade are different problems." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Both mention strict config, but one blocks loading a config while the other silently drops a field; not clearly the same bug." + }, + { + "left": "issue:47217", + "right": "issue:47429", + "accept": false, + "reason": "VL image-placeholder failure and tokenizers version upper-bound conflict are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "Same SAM3 family, but missing attribute and memory overflow are different failure modes." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading and Windows/ROCm distributed import dependency are unrelated." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "sys.modules access and xdist file race are different infrastructure bugs." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "One is model loading on transformers>=5.1.0, the other is tokenizer error masking when protobuf is absent; different code paths." + }, + { + "left": "issue:44623", + "right": "issue:46726", + "accept": false, + "reason": "Processor save-pretrained file omissions and AutoProcessor backend-missing errors are not the same bug." + }, + { + "left": "issue:47312", + "right": "issue:47455", + "accept": true, + "reason": "Both point to the same kernel package version-checking problem: one reports incompatibility with kernels 0.16, the other says the version check is wrong." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "Both are about SAM/Qwen model families, but one is a missing attribute and the other is a video-input ValueError." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "Load-balancing loss omission and LwDetr float16 AMP crash are unrelated training bugs." + }, + { + "left": "issue:44038", + "right": "issue:46506", + "accept": false, + "reason": "Qwen3-VL-Moe version bug and PPDocLayoutV3 CPU/CUDA mismatch are different model issues." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "Both mention Qwen3.5, but one is tied-weights warnings during fine-tuning and the other is flash-attention generation failure." + }, + { + "left": "issue:45561", + "right": "issue:47455", + "accept": false, + "reason": "Test utility file-race and kernel version-check logic are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45588", + "accept": false, + "reason": "Video-input regression in VL models and flash_attention s_aux crash are different layers and failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it treats the cluster as mostly heterogeneous and only elevates the 47312/47455 pair, which is the only clearly related duplicate-like pair from the titles. The soft-edge verdicts are generally cautious and do not overstate evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 49, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:39692", + "issue:42199", + "issue:42886", + "issue:43065", + "issue:43408", + "issue:43653", + "issue:43701", + "issue:43723", + "issue:43784", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43994", + "issue:44162", + "issue:44188", + "issue:44361", + "issue:44368", + "issue:44483", + "issue:44749", + "issue:45103", + "issue:45229", + "issue:45397", + "issue:45412", + "issue:45563", + "issue:45593", + "issue:45656", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46132", + "issue:46202", + "issue:46305", + "issue:46315", + "issue:46399", + "issue:46421", + "issue:46710", + "issue:46726", + "issue:46752", + "issue:46851", + "issue:46858", + "issue:47020", + "issue:47059", + "issue:47068", + "issue:47145", + "issue:47429" + ], + "soft_pairs": [ + "issue:46305|issue:46315", + "issue:29870|issue:30333", + "issue:43653|issue:44162", + "issue:45397|issue:45656", + "issue:43723|issue:46710", + "issue:42886|issue:45229", + "issue:43065|issue:43408", + "issue:45987|issue:46082", + "issue:45412|issue:46123", + "issue:43906|issue:46202", + "issue:45103|issue:46858", + "issue:43931|issue:44368", + "issue:47059|issue:47429", + "issue:45593|issue:47068", + "issue:43994|issue:46726", + "issue:44361|issue:44749", + "issue:42199|issue:46851", + "issue:44368|issue:45563", + "issue:46752|issue:47020", + "issue:46132|issue:47145", + "issue:44188|issue:46202", + "issue:43784|issue:46399", + "issue:43701|issue:44749", + "issue:43976|issue:47429", + "issue:44188|issue:46421", + "issue:44483|issue:46202", + "issue:39692|issue:46726", + "issue:44749|issue:45656" + ], + "prepared_review_unit_hash": "df7dcc4d705293ca8404f5a7eeaa2beafef8edf6c8e8350a1bc47ffabfd93fcf", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35903, + "estimated_input_tokens": 8976, + "estimated_eval_tokens": 18208 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated one-off bugs across tokenizers, generation, vision models, cache behavior, and training/inference tooling. None of the soft pairs look like the same underlying change or defect, so I would not merge any pair as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "No canonical duplicate stands out: the set is heterogeneous, and the soft pairs do not share a single concrete bug or code path.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46305 is the strongest standalone report here: it has a precise off-by-one in `sliding_window_mask_function`, a clear symptom, and an actionable fix path.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "MoonshineStreaming masking off-by-one vs tokenizer `convert_tokens_to_ids` slowdown; different subsystems and failures." + }, + { + "left": "issue:29870", + "right": "issue:30333", + "accept": false, + "reason": "Bart `generate()` UnboundLocalError vs MLflow reporting after failed training; unrelated bugs." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "BigBirdTokenizer special-token registration bug vs ESM2 model breakage; not the same code path." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma-4 ZeRO-3 `from_pretrained` loading issue vs Deepspeed optimizer stepping twice; different problems." + }, + { + "left": "issue:43723", + "right": "issue:46710", + "accept": false, + "reason": "Tokenizer loading issue in v5 vs DeepSeek accuracy regression after upgrade; distinct symptoms and fixes." + }, + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Offline tokenizer cache loading failure vs Gemma4 multi-GPU CUDA OOM." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "Dummy Conv2d in Sam3PixelDecoder vs sam3_video/sam3_tracker warning; same family, different issue." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "Cache return-type contract bug vs LlamaConfig head_dim validation; unrelated." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "RT-DETR memory not released on deletion vs MaskGenerationPipeline final-batch handling." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "Reproduction of another issue vs Gemma pipeline roundtrip failure on `\u2581`; no shared bug." + }, + { + "left": "issue:45103", + "right": "issue:46858", + "accept": false, + "reason": "auto_docstring annotation crash vs static-cache seq-length tensor arithmetic; different code paths." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch vs tie_word_embeddings warning; not the same underlying defect." + }, + { + "left": "issue:47059", + "right": "issue:47429", + "accept": false, + "reason": "CLI completion regression with huggingface_hub vs tokenizers version upper-bound packaging issue." + }, + { + "left": "issue:45593", + "right": "issue:47068", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior vs Deepspeed host-synchronization in loss computation." + }, + { + "left": "issue:43994", + "right": "issue:46726", + "accept": false, + "reason": "SigLIP2 docs/example model-processor mismatch and quantization failure vs missing-vision-backend AutoProcessor error." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "MLLukeTokenizer AttributeError on tasks vs filtering slowdown after Transformers upgrade." + }, + { + "left": "issue:42199", + "right": "issue:46851", + "accept": false, + "reason": "DETR cardinality error for no-background-class models vs Qwen3.5 variable-length `position_ids` bug." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "Tie-word-embeddings warning during fine-tuning vs stale `num_return_sequences` warning." + }, + { + "left": "issue:46752", + "right": "issue:47020", + "accept": false, + "reason": "Empty chat-template list `IndexError` vs XLMRobertaTokenizer `Unigram(vocab=...)` type mismatch." + }, + { + "left": "issue:46132", + "right": "issue:47145", + "accept": false, + "reason": "AttentionInterface registration behavior change vs FP8 checkpoint loading without weight_scale; different failures." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "`torch.compile` attention-kernel divergence vs Gemma token roundtrip bug." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "Missing `nn` import on import vs missing decorator for `apply_rotary_pos_emb` registration." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "`resume_from_checkpoint` key mismatch vs performance regression after upgrade." + }, + { + "left": "issue:43976", + "right": "issue:47429", + "accept": false, + "reason": "Python version support regression vs tokenizers upper-bound packaging constraint." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "Attention-kernel divergence under compile vs cache lazy initialization causing recompilation." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "`/v1/chat/completions` request rejection vs Gemma pipeline roundtrip failure; unrelated." + }, + { + "left": "issue:39692", + "right": "issue:46726", + "accept": false, + "reason": "SigLIP2 docs example errors vs generic missing-vision-backend processor error; not the same bug." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Filtering slowdown after upgrade vs Deepspeed double optimizer step; different subsystems." + } + ] + }, + "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": 28, + "review_unit_index": 50, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:36010", + "issue:41628", + "issue:42907", + "issue:42994", + "issue:43065", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43723", + "issue:43873", + "issue:44038", + "issue:44162", + "issue:44188", + "issue:44279", + "issue:44297", + "issue:44373", + "issue:44451", + "issue:44623", + "issue:44661", + "issue:44869", + "issue:44871", + "issue:44945", + "issue:44964", + "issue:45137", + "issue:45230", + "issue:45356", + "issue:45375", + "issue:45397", + "issue:45478", + "issue:45561", + "issue:45656", + "issue:46123", + "issue:46132", + "issue:46326", + "issue:46463", + "issue:46821", + "issue:46851", + "issue:46877", + "issue:47000", + "issue:47059", + "issue:47217", + "issue:47429", + "issue:47448" + ], + "soft_pairs": [ + "issue:42994|issue:43482", + "issue:43065|issue:46123", + "issue:43475|issue:46821", + "issue:43475|issue:43643", + "issue:43646|issue:44162", + "issue:46463|issue:46877", + "issue:44297|issue:44623", + "issue:44964|issue:45137", + "issue:42907|issue:47448", + "issue:44373|issue:46851", + "issue:28282|issue:30990", + "issue:43723|issue:46326", + "issue:44623|issue:45656", + "issue:45230|issue:45397", + "issue:44871|issue:46132", + "issue:44188|issue:46851", + "issue:46877|issue:47000", + "issue:44451|issue:47429", + "issue:45561|issue:47217", + "issue:43232|issue:43673", + "issue:44661|issue:47059", + "issue:36010|issue:41628", + "issue:44869|issue:45356", + "issue:44279|issue:47429", + "issue:43873|issue:44945", + "issue:45375|issue:45478", + "issue:44038|issue:44623" + ], + "prepared_review_unit_hash": "82ea9501d21103999cc9a86cdfd11442dd082f6a0a35f6a60cdf53138bedaf9e", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35254, + "estimated_input_tokens": 8814, + "estimated_eval_tokens": 17884 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a mixed cluster of mostly unrelated Transformers issues, many around v5 loading/init regressions, tokenizer/model-specific failures, and a few unrelated utility/test problems. None of the soft pairs are strong enough to treat as duplicates under a strict code-path standard; the broadest representative issue is the generic v5 custom model initialization regression.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43646 is the broadest and most representative issue in the set: it describes a general Transformers 5.0.0 model-initialization break, which aligns with the cluster\u2019s recurring v5 compatibility theme better than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43646 is the best issue to keep as the cluster representative because it is general, clearly user-facing, and likely to absorb the broadest class of v5 initialization regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "One is about quantized model saving, the other about GGUF loading in v5; different operations and failure modes." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "SAM3 pixel decoder internals vs MaskGenerationPipeline batch-finalization logic are unrelated code paths." + }, + { + "left": "issue:43475", + "right": "issue:46821", + "accept": false, + "reason": "Different models and different missing-attribute errors; no shared concrete bug." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "SAM3 vision encoder output attribute issue versus AutoConfig trust_remote_code field loss; distinct loading/serialization problems." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "Both are v5 regressions, but the concrete failures differ: custom model init versus ESM2-specific breakage." + }, + { + "left": "issue:46463", + "right": "issue:46877", + "accept": false, + "reason": "An incorrect decoder-input error message is not the same bug as docstring diagnostics printing at import time." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Tokenizer save metadata mismatch and processor save missing files are different save_pretrained problems." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal loading and DeepSpeed ZeRO3 deque errors are unrelated runtime failures." + }, + { + "left": "issue:42907", + "right": "issue:47448", + "accept": false, + "reason": "Missing PyTorch dependency error versus Gemma4 architecture recognition failure are different import/load issues." + }, + { + "left": "issue:44373", + "right": "issue:46851", + "accept": false, + "reason": "Wrong docstring text and variable-length position_ids runtime behavior are not the same bug." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "ImportError for missing PyTorch is unrelated to Sentence Transformers hanging during load." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "AutoTokenizer loading in v5 and continue_final_message behavior for mistral_common are separate tokenizer-path issues." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Missing processor save artifacts and duplicate optimizer steps under deepspeed are unrelated." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "A generic bug report title does not establish the same concrete failure as the Gemma-4 Zero3 loading issue." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma-3 EOS-token mismatch and AttentionInterface.register behavior change are different model/config issues." + }, + { + "left": "issue:44188", + "right": "issue:46851", + "accept": false, + "reason": "torch.compile attention-kernel divergence is a different code path than Qwen3.5 variable-length position_ids handling." + }, + { + "left": "issue:46877", + "right": "issue:47000", + "accept": false, + "reason": "Import-time doc-lint output and RTDetrLoss infeasible cost matrix are unrelated." + }, + { + "left": "issue:44451", + "right": "issue:47429", + "accept": false, + "reason": "Model loading failure and dependency upper-bound packaging issue are not the same underlying bug." + }, + { + "left": "issue:45561", + "right": "issue:47217", + "accept": false, + "reason": "pytest-xdist file-race in tests is unrelated to VL placeholder handling in user text." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Generation kwargs update after sync_gpus and missing GenerationMixin cache in chunked_prefill are different generation bugs." + }, + { + "left": "issue:44661", + "right": "issue:47059", + "accept": false, + "reason": "Tokenizer mapping-name edge case and huggingface_hub CLI completion regression are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Missing GenerationMixin import and missing AutoImageProcessor import are different API export regressions." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Whisper trailing replacement-character crash and Kimi-K2.5 codec-handling regression are not clearly the same code-path." + }, + { + "left": "issue:44279", + "right": "issue:47429", + "accept": false, + "reason": "General dependency complaint versus a specific tokenizers version-constraint problem are different issues." + }, + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Quantization offloading behavior and pipeline-parallelism output divergence are separate runtime behaviors." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": false, + "reason": "Both mention Qwen3.5 MoE, but one is a missing config field dropped by strict serialization and the other is a broader from_pretrained failure; the concrete bugs are not identical enough to merge." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe compatibility with Transformers 5.0 and processor.save_pretrained missing files 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": 28, + "review_unit_index": 51, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29925", + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:41628", + "issue:42175", + "issue:42371", + "issue:42831", + "issue:42898", + "issue:43232", + "issue:43278", + "issue:43479", + "issue:43784", + "issue:43824", + "issue:43976", + "issue:44038", + "issue:44186", + "issue:44222", + "issue:44351", + "issue:44496", + "issue:44568", + "issue:44743", + "issue:44792", + "issue:44843", + "issue:44857", + "issue:44933", + "issue:44945", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45310", + "issue:45375", + "issue:45405", + "issue:45406", + "issue:45561", + "issue:45593", + "issue:45712", + "issue:45753", + "issue:45812", + "issue:46202", + "issue:46701", + "issue:46829", + "issue:46877", + "issue:47059", + "issue:47068", + "issue:47085" + ], + "soft_pairs": [ + "issue:45081|issue:45103", + "issue:43278|issue:45137", + "issue:45310|issue:45375", + "issue:41628|issue:44933", + "issue:42371|issue:45092", + "issue:29925|issue:30333", + "issue:45406|issue:46202", + "issue:46202|issue:46701", + "issue:44857|issue:46877", + "issue:44743|issue:47068", + "issue:43232|issue:44792", + "issue:45593|issue:46877", + "issue:43784|issue:43824", + "issue:42175|issue:44933", + "issue:44038|issue:45310", + "issue:41628|issue:44351", + "issue:30333|issue:33290", + "issue:45405|issue:46877", + "issue:42898|issue:44568", + "issue:44222|issue:45310", + "issue:42175|issue:45712", + "issue:42898|issue:45812", + "issue:45561|issue:47059", + "issue:44945|issue:45753", + "issue:42831|issue:44222", + "issue:44496|issue:45561", + "issue:44843|issue:47085", + "issue:44186|issue:47085", + "issue:30990|issue:35141", + "issue:44568|issue:47085", + "issue:43479|issue:46829", + "issue:43976|issue:44933" + ], + "prepared_review_unit_hash": "fdfc324ab3aee9eaffab03a1d153de7fb4ae33e3bd1f36489d55f31211a9f30f", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37364, + "estimated_input_tokens": 9341, + "estimated_eval_tokens": 18938 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This looks like a noisy, heterogeneous cluster rather than a real duplicate set: the soft pairs span unrelated tokenizer, import, training, generation, FP8, and packaging issues, so none should be merged. If an anchor is needed, 47085 is the most substantive standalone report, but it does not unify the cluster.", + "confidence": 0.96, + "canonical_issue_reason": "No single canonical issue stands out because the items are mostly unrelated false-positive similarities across different subsystems and failure modes.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47085 is the clearest high-signal standalone bug report in the set: it has a concrete failure mode, discussion, and inbound references. It is not a cluster-wide duplicate target, just the strongest single issue here.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Different failures: Mistral tokenizer regex handling vs docstring annotation parsing. No shared code path." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "Unrelated symptoms: embedding dtype drift in evaluation vs a DeepSpeed ZeRO3 deque error." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": false, + "reason": "Both involve Qwen3.5-MoE, but one is a general from_pretrained error and the other is a missing vision-config field. Not the same bug." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Both are import-related, but they target different symbols and code paths." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "Package/backend install behavior vs meta-initialization incompatibility for old InternVL2 checkpoints." + }, + { + "left": "issue:29925", + "right": "issue:30333", + "accept": false, + "reason": "Tokenizer fast-mode enforcement and MLflow reporting are unrelated." + }, + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Gemma processor attribute crash vs Gemma pipeline roundtrip failure; different concrete defects." + }, + { + "left": "issue:46202", + "right": "issue:46701", + "accept": false, + "reason": "Both are Gemma-related, but one is token roundtrip and the other is an image-text pipeline/output-type issue." + }, + { + "left": "issue:44857", + "right": "issue:46877", + "accept": false, + "reason": "CUDA/AMP crash in a loss module is unrelated to auto_docstring import-time stdout noise." + }, + { + "left": "issue:44743", + "right": "issue:47068", + "accept": false, + "reason": "Qwen recurrent-state reset bug and a CUDA host-synchronization concern in deepspeed_sp_compute_loss are different issues." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs update bug vs Janus image generation test failure; not the same code path." + }, + { + "left": "issue:45593", + "right": "issue:46877", + "accept": false, + "reason": "D-FINE auxiliary-loss behavior is unrelated to auto_docstring diagnostics printing." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Separate import failures for different packages/models; no common underlying bug." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Backend extras packaging issue vs missing image_utils export; different failures." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen MoE models, but they are different model families and different failure surfaces." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Import errors for different public symbols; not the same underlying regression." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow failure reporting and AdaFactor OOM in DeepSpeed are unrelated." + }, + { + "left": "issue:45405", + "right": "issue:46877", + "accept": false, + "reason": "Dependency version bump/release timing issue vs docstring lint output issue." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is cleanup-space behavior and the other is add_special_tokens/BOS-EOS handling." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "FP8 save_pretrained behavior and Qwen3.5 from_pretrained loading are different code paths." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Packaging/backend extras and leftover dummy PT classes are unrelated." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Different tokenizer regressions: cleanup-space behavior vs Granite token ID generation." + }, + { + "left": "issue:45561", + "right": "issue:47059", + "accept": false, + "reason": "A pytest-xdist file race and a huggingface_hub CLI compatibility break are unrelated." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline parallelism output corruption and a Qwen3_5 image infinite loop are different failures." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "FP8 accuracy issue vs FP8 save_pretrained issue; same broad area, not the same bug." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Unrecognized model/config problem vs xdist captured_info.txt race; unrelated." + }, + { + "left": "issue:44843", + "right": "issue:47085", + "accept": false, + "reason": "Offline mistral-regex patch bug and tokenizer thread-safety regression are different code paths and symptoms." + }, + { + "left": "issue:44186", + "right": "issue:47085", + "accept": false, + "reason": "LayoutLMv2Tokenizer batching crash is not the same as the fast-tokenizer thread-safety regression." + }, + { + "left": "issue:44568", + "right": "issue:47085", + "accept": false, + "reason": "BOS/EOS insertion regression and tokenizer thread-safety are distinct issues." + }, + { + "left": "issue:43479", + "right": "issue:46829", + "accept": false, + "reason": "Config default initialization bug vs output field misassignment in VideoPrism; unrelated." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python-version compatibility break and missing image_utils import 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": 28, + "review_unit_index": 52, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:22355", + "issue:28282", + "issue:30333", + "issue:34689", + "issue:36331", + "issue:39692", + "issue:42222", + "issue:42673", + "issue:42796", + "issue:42898", + "issue:43116", + "issue:43295", + "issue:43479", + "issue:43704", + "issue:43723", + "issue:43874", + "issue:44016", + "issue:44060", + "issue:44079", + "issue:44190", + "issue:44295", + "issue:44448", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44869", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45200", + "issue:45310", + "issue:45405", + "issue:45600", + "issue:45656", + "issue:45712", + "issue:46077", + "issue:46123", + "issue:46132", + "issue:46209", + "issue:46231", + "issue:46431", + "issue:46665", + "issue:46693", + "issue:46909", + "issue:47059", + "issue:47085" + ], + "soft_pairs": [ + "issue:42673|issue:46693", + "issue:39692|issue:46909", + "issue:44060|issue:45125", + "issue:43723|issue:44568", + "issue:42673|issue:43704", + "issue:43479|issue:44079", + "issue:44623|issue:45310", + "issue:30333|issue:36331", + "issue:42898|issue:44448", + "issue:42898|issue:43295", + "issue:45600|issue:47059", + "issue:46209|issue:46431", + "issue:30333|issue:34689", + "issue:42673|issue:47085", + "issue:42898|issue:44779", + "issue:43479|issue:44869", + "issue:22355|issue:28282", + "issue:44016|issue:46231", + "issue:46077|issue:46123", + "issue:42222|issue:45200", + "issue:44734|issue:45103", + "issue:42796|issue:43723", + "issue:44448|issue:46132", + "issue:44295|issue:45656", + "issue:43116|issue:44190", + "issue:45405|issue:45712", + "issue:43874|issue:45092", + "issue:43479|issue:46665" + ], + "prepared_review_unit_hash": "b72e4cc35fb5dae4798fd4e531bf376def76643253eb1a100fabc954e88ae98a", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35723, + "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 mostly a heterogeneous mix of unrelated transformer bugs/docs regressions. The only clear duplicate-looking pair is the repeated Qwen3ForCausalLM VRAM leak report; the rest are merely same-subsystem or same-version similarity, not the same underlying issue.", + "confidence": 0.89, + "canonical_issue_reason": "issue:42673 is the best canonical issue because it has an exact duplicate in issue:43704 and describes a concrete, specific bug with a clear symptom and reproduction scope.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:42673 is the strongest issue to keep as the representative anchor: it is specific, duplicated verbatim by another report, and more concrete than the broader version/regression reports in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42673", + "right": "issue:46693", + "accept": false, + "reason": "Both mention performance/VRAM around Qwen3, but one is a dataloader-thread leak and the other is an FA2 long-context performance regression; different bug paths." + }, + { + "left": "issue:39692", + "right": "issue:46909", + "accept": false, + "reason": "Both are broken documentation examples, but they target different models and failures (SigLIP2 example vs DIA example), so not the same issue." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Both involve Qwen3 model internals, but one is a tied-weights warning and the other is a missing tensor-parallel plan; distinct code-path problems." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer/v5 regressions, but one is tokenizer loading and the other is add_special_tokens behavior for a specific model; not the same bug." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact same title and same Qwen3ForCausalLM VRAM leak symptom; these look like duplicate reports of the same bug." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "One is multimodal config initialization defaults, the other is ModelOutput key assignment when a key was previously None; unrelated internals." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "Different failures: missing files in processor.save_pretrained versus a Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting on failed training jobs is unrelated to CustomTrainer.compute_loss receiving an unexpected keyword argument." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "Both are v5 behavior regressions, but one is tokenization cleanup and the other is Pegasus output drift; different symptoms and fixes." + }, + { + "left": "issue:42898", + "right": "issue:43295", + "accept": false, + "reason": "The first is about tokenization cleanup behavior; the second is a processor/tokenizer API regression with image inputs. Not the same issue." + }, + { + "left": "issue:45600", + "right": "issue:47059", + "accept": false, + "reason": "CI breakage from removed config references is unrelated to the huggingface_hub CLI completion failure." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "FP8 inference on B200 and missing chat template are separate runtime problems with different code paths." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "MLflow job-status reporting and Llama 3.2 Vision model loading are unrelated bugs." + }, + { + "left": "issue:42673", + "right": "issue:47085", + "accept": false, + "reason": "A VRAM leak in Qwen3ForCausalLM is not the same as a broader tokenizer thread-safety regression." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but they affect different models and behaviors; not mergeable as one bug fix." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Multimodal config defaults and Whisper timestamp decoding are unrelated failures." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Different import errors: missing transformers.onnx versus missing PyTorch for AutoModel." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "A notebook syntax error and a doc parameter-name mismatch are separate documentation issues." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Bidirectional attention mask construction and final partial-batch result dropping are different pipeline/model behaviors." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Vitpose model breakage and Gemma 4 mm_token_type_ids defaulting are distinct model-family bugs." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "Serving KV-cache indexing and auto_docstring annotation handling are unrelated code paths." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Both involve tokenizer-related v5 behavior, but one is BPE marker handling for Ministral and the other is AutoTokenizer loading; different defects." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Pegasus output differences and AttentionInterface.register behavior changes are not the same underlying bug." + }, + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "A position_ids buffer/read error and a double optimizer step under deepspeed are unrelated training/runtime bugs." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "Different example scripts and different failures: empty multilabel outputs versus loading a local image dataset." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "A version pin mismatch for peft is unrelated to leftover dummy class cleanup." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-processor method and remote-code meta-initialization incompatibility are separate multimodal-loading problems." + }, + { + "left": "issue:43479", + "right": "issue:46665", + "accept": false, + "reason": "Multimodal config default initialization is unrelated to inconsistent from_pretrained behavior inside Python classes." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analyst output is grounded in the packet and stays conservative on the soft-edge matches. The only clearly duplicate-looking pair is issue:42673 and issue:43704, which share the exact title and the same Qwen3ForCausalLM VRAM leak symptom. The rest of the pairwise rejections appropriately distinguish different bug paths or failures rather than overstating similarity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 53, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:30990", + "issue:38175", + "issue:41628", + "issue:42757", + "issue:42831", + "issue:42907", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43475", + "issue:43504", + "issue:43723", + "issue:43986", + "issue:44075", + "issue:44190", + "issue:44222", + "issue:44297", + "issue:44327", + "issue:44373", + "issue:44484", + "issue:44485", + "issue:44623", + "issue:44933", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45245", + "issue:45310", + "issue:45357", + "issue:45381", + "issue:45656", + "issue:45712", + "issue:45834", + "issue:46050", + "issue:46082", + "issue:46132", + "issue:46209", + "issue:46231", + "issue:46305", + "issue:46369", + "issue:46402", + "issue:46424", + "issue:46614", + "issue:46726", + "issue:46909", + "issue:46932", + "issue:47059", + "issue:47276" + ], + "soft_pairs": [ + "issue:44222|issue:44297", + "issue:42831|issue:44623", + "issue:44297|issue:45656", + "issue:44933|issue:45712", + "issue:45245|issue:46132", + "issue:44075|issue:46231", + "issue:44222|issue:44623", + "issue:43723|issue:44977", + "issue:43232|issue:47276", + "issue:44373|issue:44485", + "issue:43504|issue:47059", + "issue:44297|issue:45310", + "issue:43122|issue:46209", + "issue:43232|issue:45834", + "issue:46614|issue:46932", + "issue:43475|issue:46305", + "issue:44327|issue:47059", + "issue:30990|issue:38175", + "issue:41628|issue:42757", + "issue:45071|issue:46082", + "issue:45092|issue:45357", + "issue:43408|issue:46402", + "issue:44190|issue:46082", + "issue:42907|issue:46050", + "issue:43232|issue:46424", + "issue:46369|issue:46909", + "issue:45381|issue:46305", + "issue:43986|issue:46726", + "issue:44222|issue:44484" + ], + "prepared_review_unit_hash": "928c426eaffaa4b2d4e421737e93252c0b43078f3891504832ee6c760c8a3915", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36068, + "estimated_input_tokens": 9017, + "estimated_eval_tokens": 18290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous with many title-level false positives. I see one plausible duplicate thread around AutoProcessor failures when a required vision backend is missing; everything else looks like distinct bugs or unrelated documentation/config issues.", + "confidence": 0.38, + "canonical_issue_reason": "No single issue cleanly represents the whole set. If one representative must be chosen, issue 46726 is the broadest and most general description of the only plausible duplicate thread (AutoProcessor/backend-missing error handling).", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46726 is the clearest, most general formulation of the AutoProcessor missing-vision-backend failure and can subsume the more specific report in 43986.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 model serialization and the other is tokenizer metadata; different bugs." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FineGrainedFP8 accuracy vs missing processor.save_pretrained files are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer save metadata and optimizer step duplication are different code paths." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Both touch import/export surfaces, but one is a nonexistent image_utils import and the other is dummy class cleanup/leakage; not the same defect." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category-count validation and AttentionInterface behavior changes are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer argument handling and a training docstring typo are not the same issue." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "FP8 model save_pretrained failure vs processor.save_pretrained missing files are different serialization problems." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading in v5 and Qwen3.5 flash-attention generation are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:47276", + "accept": false, + "reason": "Both involve synced_gpus, but one is a generation kwargs/update bug and the other is a warning message issue." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "A docstring mismatch and a RoPE implementation discussion are not the same bug." + }, + { + "left": "issue:43504", + "right": "issue:47059", + "accept": false, + "reason": "Pretrained model preset loading and a huggingface_hub CLI compatibility failure are unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save metadata mismatch and Qwen3.5 from_pretrained failure are different problems." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenization drift across versions and FP8 inference failure on B200 are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation state sync and a Kosmos2.5 OCR index error are different bugs." + }, + { + "left": "issue:46614", + "right": "issue:46932", + "accept": false, + "reason": "Stale phonemizer language state and Gemma4 quantized embedder dtype casting are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "SAM3 attribute error and Moonshine streaming mask off-by-one are different code paths." + }, + { + "left": "issue:44327", + "right": "issue:47059", + "accept": false, + "reason": "QA decode_spans bounds error and CLI completion API breakage are unrelated." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading hang and zero-probability SigLIP outputs are different issues." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different missing symbols in different packages." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "PretrainedConfig type checking and LlamaConfig head_dim validation are related only superficially." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Both involve model loading/saving, but the affected models and failure modes differ." + }, + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "sam3_video/sam3_tracker warning and missing Sam3Model text encoder weights are different SAM3 issues." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading and LlamaConfig validation are unrelated." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Both are quantization save/load problems, but dequantized Ministrals/Devstrals and sinq serialization are different mechanisms." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation sync_gpus handling and cache length behavior are unrelated." + }, + { + "left": "issue:46369", + "right": "issue:46909", + "accept": false, + "reason": "Memory overflow in sam3-litetext and a broken documentation example are unrelated." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Qwen video position_ids and Moonshine streaming mask off-by-one are different model-specific bugs." + }, + { + "left": "issue:43986", + "right": "issue:46726", + "accept": true, + "reason": "Both describe AutoProcessor failing when a required vision backend is missing; the same error-handling fix could plausibly cover both." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 model save_pretrained and max_shard_size default behavior are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted soft edge, issue:43986 vs issue:46726, is plausibly the same AutoProcessor/vision-backend error-handling problem. The other pairwise rejects are supported by the titles and do not overclaim duplicates." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 54, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28218", + "issue:39401", + "issue:41628", + "issue:42796", + "issue:42831", + "issue:42886", + "issue:42994", + "issue:43066", + "issue:43122", + "issue:43408", + "issue:43540", + "issue:43576", + "issue:43644", + "issue:43756", + "issue:43784", + "issue:43824", + "issue:43986", + "issue:44038", + "issue:44060", + "issue:44188", + "issue:44261", + "issue:44263", + "issue:44279", + "issue:44295", + "issue:44360", + "issue:44373", + "issue:44483", + "issue:44568", + "issue:44623", + "issue:44829", + "issue:44964", + "issue:45092", + "issue:45310", + "issue:45381", + "issue:45406", + "issue:45600", + "issue:45753", + "issue:46002", + "issue:46305", + "issue:46328", + "issue:46369", + "issue:46655", + "issue:46762", + "issue:46909", + "issue:47059", + "issue:47217" + ], + "soft_pairs": [ + "issue:44295|issue:44623", + "issue:44279|issue:46328", + "issue:44263|issue:45092", + "issue:44360|issue:46762", + "issue:41628|issue:43824", + "issue:44263|issue:44373", + "issue:46305|issue:46655", + "issue:22355|issue:28218", + "issue:22355|issue:27957", + "issue:43784|issue:45600", + "issue:44483|issue:45406", + "issue:43756|issue:44060", + "issue:42831|issue:45310", + "issue:44964|issue:47059", + "issue:43576|issue:43784", + "issue:44188|issue:44373", + "issue:44261|issue:45092", + "issue:42796|issue:43122", + "issue:42831|issue:44038", + "issue:44263|issue:44360", + "issue:43644|issue:46305", + "issue:43986|issue:47217", + "issue:44829|issue:46002", + "issue:43540|issue:46909", + "issue:39401|issue:44568", + "issue:42831|issue:45753", + "issue:43408|issue:46369", + "issue:45381|issue:46909", + "issue:42886|issue:43066", + "issue:42886|issue:42994" + ], + "prepared_review_unit_hash": "07765c1b7523c4bfa869cda36b82a52222b5399aca4150d5733e031367abb368", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36387, + "estimated_input_tokens": 9097, + "estimated_eval_tokens": 18450 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose mix of Transformers issues, mostly tokenizer/version-regression reports plus several unrelated model, import, and processor bugs. None of the soft pairs look like safe duplicates, so they should stay separate.", + "confidence": 0.84, + "canonical_issue_reason": "issue:43122 is the best available anchor because it is the broadest tokenizer regression report (same tokenizer, different behavior across 4.57.3 to 5.0) and sits near several other tokenizer-related issues. That said, the cluster is still too heterogeneous for a single true duplicate canon.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43122 is the strongest representative issue among this set because it captures the common tokenizer/version-regression theme better than the more model-specific reports. It is still only a partial fit for the overall cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "Buffer read error and processor.save_pretrained missing files are different failure modes in different code paths." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A dependency issue and a padding performance regression are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split return-value handling in GlmMoeDsaIndexer is unrelated to remote-code meta-initialization incompatibility." + }, + { + "left": "issue:44360", + "right": "issue:46762", + "accept": false, + "reason": "A missing ReLU in the DSA indexer and Minimax attention numerical mismatch are different bugs." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but for different symbols and model paths; not the same underlying missing export." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "One is a runtime indexing bug, the other is a docstring problem." + }, + { + "left": "issue:46305", + "right": "issue:46655", + "accept": false, + "reason": "Sliding-window off-by-one and video padding calculation are separate processor bugs." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "Missing transformers.onnx module import and tokenizer spacing behavior are unrelated." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Missing transformers.onnx module and XLMRoberta Flash Attention 2 are unrelated issues." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Import-time NameError for nn and stale auto_mappings references are different breakages." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "A chat-completions API rejection and a missing _tokenizer attribute in Gemma4Processor are different failures." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warning are separate model-architecture bugs." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy drift and Qwen3.5 Moe from_pretrained error are not the same bug." + }, + { + "left": "issue:44964", + "right": "issue:47059", + "accept": false, + "reason": "Phi-4 multimodal loading and a CLI incompatibility with huggingface_hub are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Broken transformers env CLI and a missing nn import are different surfaces and causes." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "torch.compile attention-kernel divergence and a wrong docstring are unrelated." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Missing rms_norm_eps in MLA q_a_layernorm and old InternVL2 meta-init incompatibility are different issues." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Both involve tokenization regressions, but one is model-specific BPE-marker handling while the other is a broader version-to-version behavior change." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "FineGrainedFP8 accuracy issues and Qwen3-VL-Moe loading/behavior problems are not the same underlying defect." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Two different DSA-indexer bugs: one about torch.split outputs, one about a missing ReLU." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer junk filling and a sliding-window mask off-by-one are separate regressions." + }, + { + "left": "issue:43986", + "right": "issue:47217", + "accept": false, + "reason": "A torchvision-missing crash in AutoProcessor and VL failures when the placeholder is in user text have different triggers." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "FlashAttention-3 degenerate training and NaNs after checkout+init are different numerical/training problems." + }, + { + "left": "issue:43540", + "right": "issue:46909", + "accept": false, + "reason": "Qwen3OmniMoe video-input ValueError and a broken DIA documentation example are unrelated." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Wrong offset_mapping and missing BOS/EOS with add_special_tokens are distinct tokenizer bugs." + }, + { + "left": "issue:42831", + "right": "issue:45753", + "accept": false, + "reason": "FP8 accuracy drift and Qwen3_5 infinite looping on an image are different failure modes." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "sam3_video vs sam3_tracker warning and sam3-litetext memory overflow are different sam3-family issues." + }, + { + "left": "issue:45381", + "right": "issue:46909", + "accept": false, + "reason": "Incorrect vision_position_ids in Qwen2.5-VL and a DIA docs bug are unrelated." + }, + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure and wrong tokenizer decoder type are both v5 tokenizer issues, but not the same code-path problem." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Tokenizer offline-cache loading and quantized model saving 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": 28, + "review_unit_index": 55, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28348", + "issue:29159", + "issue:30990", + "issue:33357", + "issue:39379", + "issue:39401", + "issue:41628", + "issue:42371", + "issue:42757", + "issue:42907", + "issue:43116", + "issue:43122", + "issue:43329", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43646", + "issue:43701", + "issue:43824", + "issue:43906", + "issue:43976", + "issue:44038", + "issue:44117", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44351", + "issue:44488", + "issue:44521", + "issue:44625", + "issue:44704", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:44995", + "issue:45200", + "issue:45216", + "issue:45278", + "issue:45290", + "issue:45478", + "issue:46002", + "issue:46172", + "issue:46326", + "issue:46640", + "issue:46697", + "issue:47059" + ], + "soft_pairs": [ + "issue:43646|issue:46002", + "issue:43329|issue:47059", + "issue:43122|issue:46326", + "issue:27957|issue:30990", + "issue:30990|issue:33357", + "issue:42757|issue:45278", + "issue:44625|issue:44829", + "issue:39379|issue:46172", + "issue:22355|issue:29159", + "issue:39401|issue:44521", + "issue:42907|issue:44964", + "issue:39401|issue:44488", + "issue:44117|issue:44704", + "issue:44295|issue:44297", + "issue:41628|issue:45278", + "issue:43122|issue:44521", + "issue:42907|issue:44945", + "issue:43976|issue:44945", + "issue:44704|issue:45290", + "issue:42907|issue:43452", + "issue:43408|issue:43475", + "issue:43824|issue:44351", + "issue:42371|issue:44263", + "issue:43116|issue:45200", + "issue:44038|issue:44297", + "issue:43122|issue:43906", + "issue:45216|issue:45478", + "issue:28348|issue:46640", + "issue:44995|issue:46697", + "issue:43701|issue:44295" + ], + "prepared_review_unit_hash": "edb5ed92be789fac0841be4b3379d47fd33059ac8c60b8d475ca96ead22990c8", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36297, + "estimated_input_tokens": 9075, + "estimated_eval_tokens": 18406 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous. The soft pairs share broad vocabulary (tokenizer, import errors, flash attention, multimodal models), but they do not describe the same underlying bug/change closely enough to merge as duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45278 is the broadest user-facing regression report here (multiple import errors after upgrading), so it is the closest thing to a cluster anchor, though the cluster itself is not a true duplicate set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 has the clearest broad regression framing and is the most likely to represent the common failure mode among the mixed import/version-related reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Different failures: custom model init breakage vs NaNs in classification-head init/checkout. Same era, not the same code-path bug." + }, + { + "left": "issue:43329", + "right": "issue:47059", + "accept": false, + "reason": "Unrelated subsystems: multimodal token counting bug vs CLI completion attribute breakage from huggingface_hub." + }, + { + "left": "issue:43122", + "right": "issue:46326", + "accept": false, + "reason": "Both tokenizer-related, but one is version-to-version tokenization drift and the other is continue_final_message behavior for mistral_common. Not the same defect." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 support request for XLM-Roberta vs a loading hang in Sentence Transformers. Different models and symptoms." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "Loading hang vs Mac bus error on a CLIP community model. No shared concrete bug path." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "Both mention imports, but one is a specific missing symbol from huggingface_hub and the other is a broad upgrade regression. Not the same issue." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "Qwen3.5 label propagation bug vs flash_attention_3 training degeneration. Different model behavior and code paths." + }, + { + "left": "issue:39379", + "right": "issue:46172", + "accept": false, + "reason": "Separate multimodal/math bugs in different models; no shared failing function or fix." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "Missing transformers.onnx module import vs slow/fast tokenizer inconsistency. Different components and failure modes." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Both concern tokenization, but one is offset_mapping for Qwen3 and the other is assistant masks in multimodal chat templating." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Saving dequantized checkpoints vs loading Phi-4 multimodal. Different model families and unrelated code paths." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Wrong offset mapping in Qwen3 tokenizer vs failure to load a specific BERT model. Not the same defect." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Tokenizer mapping fallback bug vs AutoProcessor kwargs propagation. Different APIs and root causes." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Reading a buffer-registered position_ids error vs tokenizer_class mismatch on save_pretrained. Different serialization/state issues." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Import name error for AutoImageProcessor vs a broad post-upgrade import breakage. Related only at a high level." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer drift across versions vs multimodal assistant mask generation. Same broad area, different bugs." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Checkpoint save/dequantization failure vs incorrect pipeline-parallel inference output. No shared concrete fix." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python-version compatibility regression vs pipeline-parallel output corruption. Different symptoms and code paths." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "Processor metadata/download kwargs bug vs chat_template crash on tool-call messages. Distinct functionality." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Dequantized save path vs gguf/from_pretrained loading breakage. Different model-loading workflows." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "SAM3 model-type warning vs missing attribute in SAM3 video encoder output. Related model family, but not the same bug." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Different missing imports: Qwen2.5-VL class vs HybridCache. Same broad import category, not one duplicate." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 API warning vs a torch.split return-value issue in GlmMoeDsaIndexer. Unrelated." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example script multi-label output bug vs Gemma 4 mm_token_type_ids defaulting. Different failure points." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe compatibility issue vs tokenizer save_pretrained metadata mismatch. Different problems." + }, + { + "left": "issue:43122", + "right": "issue:43906", + "accept": false, + "reason": "Tokenizer behavior change across versions vs isolated reproduction of another issue. No concrete overlap." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5 and version 5.4.0, but one is a bad saved checkpoint and the other is a from_pretrained loading error. Related area, not clearly the same bug." + }, + { + "left": "issue:28348", + "right": "issue:46640", + "accept": false, + "reason": "Specific GPT2 Flash Attention 2 support request vs a generic flash-attention issue report. Not the same code path." + }, + { + "left": "issue:44995", + "right": "issue:46697", + "accept": false, + "reason": "GlmMoeDsa stale cache on second forward pass vs Swin2SR image padding overflow. Completely different models." + }, + { + "left": "issue:43701", + "right": "issue:44295", + "accept": false, + "reason": "resume_from_checkpoint key mismatch vs position_ids buffer access error. Different state-loading 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": 28, + "review_unit_index": 56, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:22355", + "issue:29127", + "issue:36296", + "issue:38617", + "issue:39379", + "issue:39401", + "issue:42175", + "issue:42222", + "issue:42673", + "issue:42757", + "issue:42898", + "issue:43116", + "issue:43232", + "issue:43278", + "issue:43493", + "issue:43611", + "issue:43976", + "issue:44038", + "issue:44075", + "issue:44188", + "issue:44222", + "issue:44351", + "issue:44360", + "issue:44483", + "issue:44485", + "issue:44610", + "issue:44625", + "issue:44936", + "issue:44938", + "issue:44945", + "issue:45072", + "issue:45083", + "issue:45278", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45753", + "issue:46002", + "issue:46172", + "issue:46231", + "issue:46328", + "issue:46614", + "issue:46762", + "issue:46877", + "issue:46909", + "issue:47217", + "issue:47312" + ], + "soft_pairs": [ + "issue:43278|issue:44945", + "issue:22355|issue:29127", + "issue:44038|issue:44222", + "issue:44360|issue:44485", + "issue:42175|issue:44938", + "issue:46172|issue:46762", + "issue:43493|issue:45753", + "issue:42222|issue:47312", + "issue:39379|issue:43232", + "issue:39401|issue:46614", + "issue:42673|issue:45072", + "issue:44188|issue:46328", + "issue:45357|issue:45478", + "issue:42757|issue:44351", + "issue:36296|issue:43116", + "issue:46231|issue:46877", + "issue:44483|issue:47217", + "issue:44625|issue:46002", + "issue:44610|issue:46909", + "issue:42898|issue:46614", + "issue:45083|issue:45446", + "issue:43976|issue:45446", + "issue:43611|issue:46002", + "issue:38617|issue:45278", + "issue:43116|issue:43493", + "issue:44075|issue:44936" + ], + "prepared_review_unit_hash": "2f9dabe17227011267e7dec9597bb262e53193e3a6366c2413a030ae018cba09", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 34327, + "estimated_input_tokens": 8582, + "estimated_eval_tokens": 17420 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab bag of unrelated Hugging Face Transformers issues. Several pairs share a theme or model family, but none look like the same concrete bug/change closely enough to deduplicate, so no soft edge should be accepted.", + "confidence": 0.91, + "canonical_issue_reason": "issue:39401 is a clean, specific bug report with a narrowly defined symptom (wrong offset_mapping) and clear reproduction context, making it a strong representative issue among the set.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:39401 is the strongest standalone issue: concrete, model-specific, and directly actionable without being overly broad or speculative.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43278", + "right": "issue:44945", + "accept": false, + "reason": "Both involve training/evaluation behavior, but one is a dtype flip and the other is pipeline-parallel output correctness; different code paths and symptoms." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Completely different failures: missing transformers.onnx module vs a LayoutLMv3 error-message clarity problem." + }, + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Both mention MoE-related regressions, but one is a Qwen3-VL-Moe loading bug and the other is FP8 save_pretrained behavior; not the same concrete bug." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "Unrelated technical areas: DSA indexer ReLU discussion vs GLM-5 RoPE implementation." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Both are import/platform issues, but one is an extras/backend dependency omission and the other is a Python 3.14 load failure; different root causes." + }, + { + "left": "issue:46172", + "right": "issue:46762", + "accept": false, + "reason": "Both concern numerical behavior, but they affect different models and mechanisms; no evidence of the same underlying bug." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "Different model families and symptoms: SigLIP2 implementation discrepancy vs Qwen3_5 infinite-loop generation on a specific image." + }, + { + "left": "issue:42222", + "right": "issue:47312", + "accept": false, + "reason": "VitPose model breakage and a kernels-version compatibility issue are unrelated." + }, + { + "left": "issue:39379", + "right": "issue:43232", + "accept": false, + "reason": "One is an mllama masking bug, the other is generation kwargs after sync_gpus; different code paths." + }, + { + "left": "issue:39401", + "right": "issue:46614", + "accept": false, + "reason": "Tokenizer offset mapping and Wav2Vec2 phonemizer backend caching are distinct tokenizer/phonemizer bugs." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leakage in multi-threaded Qwen3ForCausalLM use is unrelated to dtype mismatches in inference." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "Attention-kernel branching under torch.compile is unrelated to SequenceFeatureExtractor.pad performance overhead." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5, but one is incorrect save_pretrained keys and the other is a from_pretrained import/load error in newer Transformers." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols from different packages; no shared underlying bug is established." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor-parallel training bug and multi-label classification example returning empty results are unrelated." + }, + { + "left": "issue:46231", + "right": "issue:46877", + "accept": false, + "reason": "Documentation parameter mismatch is not the same as auto_docstring printing diagnostics at import time." + }, + { + "left": "issue:44483", + "right": "issue:47217", + "accept": false, + "reason": "The chat/completions request rejection issue and the VL image-placeholder failure are different API/model behaviors." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "Config propagation for Qwen3.5 num_labels and NaNs in classification heads are different bugs with different failure modes." + }, + { + "left": "issue:44610", + "right": "issue:46909", + "accept": false, + "reason": "Processor input-size mismatch and a broken DIA docs example are not the same issue." + }, + { + "left": "issue:42898", + "right": "issue:46614", + "accept": false, + "reason": "Tokenization cleanup semantics and phonemizer backend state retention are unrelated tokenizer/audio bugs." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "A feature-length calculation bug in qwen3_omni_moe is unrelated to a PyTorch version-check issue in flex_attention." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python-version compatibility and PyTorch version gating are separate environment-compatibility problems." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "base_model_prefix loading regression and NaNs in classification heads are different failures." + }, + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing import name and a broad set of import errors after upgrade may be related at a high level, but not enough to treat as the same concrete bug." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Example-script classification output bug and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling and trainer.evaluate() failing after train() are separate training-loop 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": 28, + "review_unit_index": 57, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:33453", + "issue:39379", + "issue:39401", + "issue:42371", + "issue:42617", + "issue:42673", + "issue:42796", + "issue:43064", + "issue:43066", + "issue:43116", + "issue:43208", + "issue:43408", + "issue:43618", + "issue:43673", + "issue:43688", + "issue:43742", + "issue:43824", + "issue:43825", + "issue:43827", + "issue:43874", + "issue:43937", + "issue:43992", + "issue:44117", + "issue:44291", + "issue:44355", + "issue:44466", + "issue:44485", + "issue:44704", + "issue:44779", + "issue:44977", + "issue:45071", + "issue:45081", + "issue:45341", + "issue:45362", + "issue:45405", + "issue:45446", + "issue:45478", + "issue:45563", + "issue:45588", + "issue:46002", + "issue:46231", + "issue:46328", + "issue:46710", + "issue:46726", + "issue:46829", + "issue:46894", + "issue:46981", + "issue:47116" + ], + "soft_pairs": [ + "issue:43066|issue:44779", + "issue:45362|issue:45478", + "issue:42673|issue:42796", + "issue:43937|issue:44485", + "issue:43066|issue:46710", + "issue:46726|issue:46981", + "issue:43618|issue:46829", + "issue:43116|issue:43825", + "issue:44355|issue:46231", + "issue:43827|issue:45446", + "issue:42371|issue:43874", + "issue:43992|issue:44704", + "issue:44291|issue:45071", + "issue:44117|issue:45588", + "issue:45405|issue:45446", + "issue:39379|issue:43688", + "issue:43824|issue:46981", + "issue:44117|issue:45081", + "issue:42617|issue:45341", + "issue:43673|issue:46328", + "issue:46328|issue:47116", + "issue:33453|issue:39401", + "issue:43064|issue:43673", + "issue:43208|issue:46894", + "issue:43937|issue:45563", + "issue:43408|issue:43742", + "issue:44466|issue:46002", + "issue:43937|issue:44977" + ], + "prepared_review_unit_hash": "4c73b4a067e89f1550684077b77f406968ef30c59d0966f40df259899d8b6c42", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36003, + "estimated_input_tokens": 9001, + "estimated_eval_tokens": 18258 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose thematic group of unrelated bug reports, with only broad overlap around tokenizer/v5 regressions. None of the soft pairs look like the same concrete underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.42, + "canonical_issue_reason": "issue:43066 is the most generally useful representative because it is a concrete tokenizer-v5 regression rather than a highly model-specific symptom, and several nearby items are tokenizer-related but not identical.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43066 is the strongest standalone issue in the set: specific enough to be actionable, but broad enough to serve as the clearest representative of the tokenizer-regression subset.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer-related v5 regressions, but one is about decoder type selection and the other about incorrect DeepSeek tokenization results; different failure modes and likely different fixes." + }, + { + "left": "issue:45362", + "right": "issue:45478", + "accept": false, + "reason": "Same model family, but one is a chat crash and the other is a from_pretrained error; different code paths and symptoms." + }, + { + "left": "issue:42673", + "right": "issue:42796", + "accept": false, + "reason": "Completely different bugs: VRAM leak in dataloader threads vs tokenizer BPE marker handling." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GLM-5 appears in both, but generation config validation and RoPE implementation are unrelated code paths." + }, + { + "left": "issue:43066", + "right": "issue:46710", + "accept": false, + "reason": "Both are v5/DeepSeek-adjacent regressions, but one is a tokenizer decoder-type issue and the other is an accuracy regression; not the same bug." + }, + { + "left": "issue:46726", + "right": "issue:46981", + "accept": false, + "reason": "AutoProcessor backend-missing handling is unrelated to HTTP 429 retry behavior." + }, + { + "left": "issue:43618", + "right": "issue:46829", + "accept": false, + "reason": "Both concern model output fields, but CLIP attentions and VideoPrism hidden_states are distinct output-shape/assignment bugs." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "One is a multi-label classification example bug; the other is an incorrect pipeline error message. Different subsystems and fixes." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Running compiled Python files is unrelated to a documentation parameter-name mismatch." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() and a flex_attention PyTorch version check are unrelated." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 API guidance and a GLM46V image-patch attribute error are unrelated." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Missing encoder weights on load vs AutoProcessor kwargs propagation are different loader bugs." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Both are v5-related type/config issues, but they affect different APIs and failure mechanisms." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping fallback and flash_attention s_aux handling are unrelated." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "A dependency-version bump problem is not the same as an incorrect PyTorch version gate." + }, + { + "left": "issue:39379", + "right": "issue:43688", + "accept": false, + "reason": "Different model families and different logic: masked-out mask handling vs auxiliary-loss normalization." + }, + { + "left": "issue:43824", + "right": "issue:46981", + "accept": false, + "reason": "ImportError for a missing model class and repeated 429 retries are unrelated." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Both involve tokenizer loading, but one is mapping-name None handling and the other is a Mistral regex/backend_tokenizer crash; not the same fix." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "A runtime failure in 3d_parallel.py is not the same as a small testing_utils bug." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Generation cache handling and SequenceFeatureExtractor.pad performance are unrelated." + }, + { + "left": "issue:46328", + "right": "issue:47116", + "accept": false, + "reason": "Padding performance and streaming pipeline regression are different behaviors in different layers." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Both are tokenizer-loading related, but one is a broad regression report and the other is a specific Qwen3 offset_mapping bug; not enough evidence they are the same underlying issue." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Training optimizer-state corruption on nonzero ranks is unrelated to chunked_prefill cache absence." + }, + { + "left": "issue:43208", + "right": "issue:46894", + "accept": false, + "reason": "Both mention training bugs, but xLSTM's multiple model issues and CohereASR's double-shift loss bug are different concrete problems." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GenerationConfig validation and a stale warning for num_return_sequences are different issues." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "A model-type warning and a MobileLLM loading key error are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of lm_head.weight depends on device; classification head NaNs after checkout/init are a different failure mode." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Both touch GLM/Qwen-style generation, but one is a GenerationConfig validity error and the other is flash-attention generation failure; different code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 58, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:38617", + "issue:39379", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42371", + "issue:42831", + "issue:43097", + "issue:43122", + "issue:43441", + "issue:43450", + "issue:43937", + "issue:43986", + "issue:44220", + "issue:44263", + "issue:44380", + "issue:44410", + "issue:44483", + "issue:44493", + "issue:44496", + "issue:44704", + "issue:44779", + "issue:44829", + "issue:44877", + "issue:44964", + "issue:45071", + "issue:45092", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45341", + "issue:45440", + "issue:45464", + "issue:45561", + "issue:45632", + "issue:46077", + "issue:46231", + "issue:46298", + "issue:46328", + "issue:46512", + "issue:46710", + "issue:46752", + "issue:46762", + "issue:46897", + "issue:46981", + "issue:47000", + "issue:47059", + "issue:47110" + ], + "soft_pairs": [ + "issue:42222|issue:47000", + "issue:43450|issue:43986", + "issue:44380|issue:46077", + "issue:45071|issue:46512", + "issue:45092|issue:46762", + "issue:42222|issue:43441", + "issue:39379|issue:46897", + "issue:42222|issue:46752", + "issue:44877|issue:46231", + "issue:45632|issue:46981", + "issue:39401|issue:44779", + "issue:44263|issue:46762", + "issue:44483|issue:45464", + "issue:44829|issue:45216", + "issue:45341|issue:45561", + "issue:44410|issue:45310", + "issue:45071|issue:47059", + "issue:45440|issue:46328", + "issue:39401|issue:46710", + "issue:38617|issue:44220", + "issue:43937|issue:44496", + "issue:43122|issue:46298", + "issue:44704|issue:47110", + "issue:38617|issue:41628", + "issue:44964|issue:45278", + "issue:43097|issue:44493", + "issue:42371|issue:42831" + ], + "prepared_review_unit_hash": "aa880f070434e00f8d8ac3f21c431b075c45685de315d3427c6393ed891814d0", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35562, + "estimated_input_tokens": 8891, + "estimated_eval_tokens": 18038 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No clear duplicate cluster emerges here: the soft pairs are mostly broad-similarity matches across different models, APIs, or failure modes, so they should not be merged as the same underlying bug. I would treat these as unrelated issues with superficial overlap.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42222", + "right": "issue:47000", + "accept": false, + "reason": "Both mention model/runtime failures, but one is a VitPose import breakage and the other is an RTDetr loss crash; different code paths and symptoms." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Both involve video processing, but one is an incorrect batched output shape and the other is an AutoProcessor crash without torchvision; different underlying bugs." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "Both are attention-related, but one is GPT2 scaling ignored in SDPA/FlashAttention and the other is Gemma4 building the wrong mask type; not the same defect." + }, + { + "left": "issue:45071", + "right": "issue:46512", + "accept": false, + "reason": "One is a type-checking regression in v5.4.0, the other is dtype metadata staying stale after from_config; unrelated config/model initialization issues." + }, + { + "left": "issue:45092", + "right": "issue:46762", + "accept": false, + "reason": "Both are model-behavior regressions, but one is meta-initialization incompatibility and the other is numerical mismatch in Minimax MSA attention; different failure mechanisms." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "VitPose model breakage and Ministral FlashAttention failure are separate model-specific regressions with different stacks." + }, + { + "left": "issue:39379", + "right": "issue:46897", + "accept": false, + "reason": "Both are loss/masking-related, but one is mllama mask handling and the other is Florence2 training-loss double shift; not the same bug." + }, + { + "left": "issue:42222", + "right": "issue:46752", + "accept": false, + "reason": "VitPose import/model breakage is unrelated to apply_chat_template crashing on an empty list." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "One is strict config loading for granite_speech, the other is a documentation parameter-name mismatch; no shared code-path bug." + }, + { + "left": "issue:45632", + "right": "issue:46981", + "accept": false, + "reason": "Both involve loading/caching, but one is a local trust_remote_code cache collision and the other is repeated 429 retries in AutoProcessor; different causes." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer issues, but one is wrong offset_mapping and the other is different tokenization output in v5; same area, different bug reports." + }, + { + "left": "issue:44263", + "right": "issue:46762", + "accept": false, + "reason": "A torch.split indexing/value issue is not the same as a Minimax attention numerical discrepancy." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "One is chat/completions request validation in v5.3, the other is streaming inference failure for Qwen3.5-0.8B; different API layers and failure modes." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "One is a flash_attention_3 training degeneration issue, the other is incorrect save_pretrained output for Qwen3.5; different behaviors and code paths." + }, + { + "left": "issue:45341", + "right": "issue:45561", + "accept": false, + "reason": "A small testing_utils bug is unrelated to pytest-xdist races in captured_info.txt." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Both concern Qwen3.5/MoE, but one is missing projections in layer 0 and the other is from_pretrained loading error; not the same concrete bug." + }, + { + "left": "issue:45071", + "right": "issue:47059", + "accept": false, + "reason": "A PretrainedConfig type-checking regression is unrelated to a cli test failure from huggingface_hub 1.22.0 attribute removal." + }, + { + "left": "issue:45440", + "right": "issue:46328", + "accept": false, + "reason": "DeepSeekV3MoE divergence and SequenceFeatureExtractor.pad inefficiency are entirely different issues." + }, + { + "left": "issue:39401", + "right": "issue:46710", + "accept": false, + "reason": "Both are tokenization/model-output regressions, but the reported effects and affected models differ; no evidence of a single underlying bug." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "ImportError for layer_type_validation and _torch_extract_fbank_features behavior are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "GLM generation config validation and OLMo-Hybrid model-type recognition are separate config-validation problems." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Different tokenization across versions and a noisy logging message are not the same bug." + }, + { + "left": "issue:44704", + "right": "issue:47110", + "accept": false, + "reason": "AutoProcessor kwargs forwarding and special_tokens list-vs-dict handling involve different APIs and error causes." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they refer to different missing symbols and different modules." + }, + { + "left": "issue:44964", + "right": "issue:45278", + "accept": false, + "reason": "Both are broad version-upgrade regressions, but one is a specific multimodal model load failure and the other is many import errors after upgrade; too generic to merge as one bug." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "One is removal of a config attribute without notice, the other is unexpected position_id keys across models; different regression classes." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 settings API behavior and FineGrainedFP8 accuracy are unrelated numeric/configuration concerns." + } + ] + }, + "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": 28, + "review_unit_index": 59, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:27957", + "issue:29159", + "issue:29925", + "issue:30947", + "issue:42371", + "issue:43012", + "issue:43066", + "issue:43097", + "issue:43122", + "issue:43208", + "issue:43352", + "issue:43441", + "issue:43452", + "issue:43531", + "issue:43577", + "issue:43673", + "issue:43874", + "issue:43931", + "issue:43957", + "issue:43986", + "issue:44038", + "issue:44075", + "issue:44077", + "issue:44117", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44351", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44829", + "issue:44987", + "issue:45071", + "issue:45310", + "issue:45341", + "issue:45372", + "issue:45478", + "issue:45542", + "issue:45588", + "issue:45656", + "issue:46050", + "issue:46752", + "issue:46762", + "issue:46897", + "issue:46981", + "issue:47116", + "issue:47448" + ], + "soft_pairs": [ + "issue:42371|issue:44261", + "issue:44261|issue:46762", + "issue:44077|issue:46981", + "issue:43931|issue:44410", + "issue:43208|issue:46897", + "issue:44038|issue:44230", + "issue:43066|issue:43122", + "issue:44484|issue:45310", + "issue:44075|issue:45656", + "issue:44410|issue:45478", + "issue:29925|issue:30947", + "issue:43673|issue:45341", + "issue:43097|issue:43452", + "issue:44263|issue:44485", + "issue:43931|issue:44038", + "issue:43957|issue:44351", + "issue:45071|issue:45542", + "issue:45341|issue:47448", + "issue:43441|issue:44829", + "issue:43012|issue:43577", + "issue:43531|issue:45341", + "issue:43874|issue:44261", + "issue:44117|issue:44987", + "issue:46752|issue:47116", + "issue:43097|issue:45588", + "issue:27957|issue:43352", + "issue:43986|issue:45372", + "issue:29159|issue:30947", + "issue:43097|issue:46050" + ], + "prepared_review_unit_hash": "21d76cee8e50e6375fb331cb7b645a00a80b0d056af0239aa798ee988cfcc04d", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36102, + "estimated_input_tokens": 9026, + "estimated_eval_tokens": 18308 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a mixed cluster with many unrelated Transformers issues. The only visible common thread is a loose tokenizer-regression theme, but none of the soft pairs look like the same concrete bug/change, so I would not merge any of them.", + "confidence": 0.77, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:43122 is the broadest tokenizer-regression report and the best representative of the small tokenizer-related subset, but the cluster as a whole is heterogeneous.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "Both are about precision behavior, but TF32 API guidance and a missing rms_norm_eps in MLA are different problems and fixes." + }, + { + "left": "issue:44261", + "right": "issue:46762", + "accept": false, + "reason": "Both mention numerical differences, but they affect different models/code paths and do not look like one underlying bug." + }, + { + "left": "issue:44077", + "right": "issue:46981", + "accept": false, + "reason": "SGD args not being used and repeated 429 retries in AutoProcessor are unrelated subsystems and failures." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Both involve Qwen3-family loading, but weight-shape mismatch and missing projections are different concrete defects." + }, + { + "left": "issue:43208", + "right": "issue:46897", + "accept": false, + "reason": "xLSTM training blockers and a Florence2 double-shift loss bug are unrelated." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both mention Qwen3-VL/MoE and fp8/FlashAttention, but one is a bug report and the other is fp8 inference support; not the same fix." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "These are both tokenizer regressions in v5, but one is about decoder type metadata and the other is a broad tokenization-change report; not clearly the same concrete bug." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "max_shard_size documentation/question and a Qwen3.5 MoE from_pretrained error are unrelated." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "Unused SGD args and double optimizer stepping with deepspeed are different optimizer bugs." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5/Qwen3-next style model issues, but they concern different loading/model-definition failures." + }, + { + "left": "issue:29925", + "right": "issue:30947", + "accept": false, + "reason": "Both are tokenizer-related, but SigLIP use_fast enforcement and LLaMA3 special-token behavior are separate model-specific issues." + }, + { + "left": "issue:43673", + "right": "issue:45341", + "accept": false, + "reason": "Generation cache/chunked_prefill and a testing_utils bug have no shared underlying code path." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Tie-embeddings API removal and gguf_file loading breakage are unrelated." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "A torch.split return-value bug and GLM-5 RoPE implementation are different defects in different components." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Both reference Qwen3-VL/MoE, but one is a weight-shape mismatch and the other is a different from_pretrained/model-loading issue." + }, + { + "left": "issue:43957", + "right": "issue:44351", + "accept": false, + "reason": "Meta-device loading failures and a missing HybridCache import are unrelated." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "PretrainedConfig type checking and a TensorBoard-without-TF backend error are unrelated." + }, + { + "left": "issue:45341", + "right": "issue:47448", + "accept": false, + "reason": "A testing_utils bug and Gemma4 architecture recognition are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Both mention FlashAttention, but one is an unsupported/failed model path and the other is a degenerate-training quality bug; not the same issue." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "A PyTorch precision warning and a model dtype propagation bug are different problems." + }, + { + "left": "issue:43531", + "right": "issue:45341", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and a testing_utils bug are unrelated." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing image-patch counting and a normalization-epsilon precision issue are unrelated multimodal/model bugs." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "A tokenizer mapping None assumption and a specific loading failure for physical-intelligence/fast are not clearly the same bug." + }, + { + "left": "issue:46752", + "right": "issue:47116", + "accept": false, + "reason": "Empty-list chat template crash and streaming pipeline regression are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "Tie-embeddings removal and a flash_attention.py sink-less crash are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "Both are FlashAttention 2 related, but they concern different models and one is unsupported-model gating while the other is a specific XLM-Roberta path." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "AutoProcessor needing torchvision and a mistral_common ReasoningEffort import error are different processor-loading failures." + }, + { + "left": "issue:29159", + "right": "issue:30947", + "accept": false, + "reason": "Generic slow/fast tokenizer inconsistency and LLaMA3 special-token behavior are related only at a high level, not the same bug." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "Both discuss model math correctness, but one is a missing config epsilon and the other is RoPE implementation details; not the same fix." + }, + { + "left": "issue:45341", + "right": "issue:47448", + "accept": false, + "reason": "A testing utility issue and a Gemma4 architecture detection bug are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Embedding-tie API removal and sinq save/load quantization 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": 28, + "review_unit_index": 60, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:30947", + "issue:33453", + "issue:34567", + "issue:38617", + "issue:41950", + "issue:42757", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43097", + "issue:43208", + "issue:43550", + "issue:43673", + "issue:43873", + "issue:43874", + "issue:43937", + "issue:43986", + "issue:44062", + "issue:44117", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44315", + "issue:44360", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44704", + "issue:44936", + "issue:45103", + "issue:45341", + "issue:45362", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45753", + "issue:46123", + "issue:46172", + "issue:46298", + "issue:46365", + "issue:46369", + "issue:46655", + "issue:46664", + "issue:46701", + "issue:46710", + "issue:46752", + "issue:46821", + "issue:46981" + ], + "soft_pairs": [ + "issue:44263|issue:46172", + "issue:44360|issue:46172", + "issue:44484|issue:45478", + "issue:43097|issue:43873", + "issue:43208|issue:43550", + "issue:44261|issue:44485", + "issue:45440|issue:46172", + "issue:42907|issue:43937", + "issue:45753|issue:46369", + "issue:43986|issue:45341", + "issue:44246|issue:46981", + "issue:44117|issue:46701", + "issue:43012|issue:46821", + "issue:30947|issue:34567", + "issue:44062|issue:44117", + "issue:44410|issue:45375", + "issue:43673|issue:46123", + "issue:30947|issue:33453", + "issue:43874|issue:44263", + "issue:46172|issue:46710", + "issue:44410|issue:45362", + "issue:44936|issue:46365", + "issue:46298|issue:46752", + "issue:46655|issue:46664", + "issue:44704|issue:45103", + "issue:43986|issue:44484", + "issue:38617|issue:41950", + "issue:38617|issue:42757", + "issue:43064|issue:44315" + ], + "prepared_review_unit_hash": "fabba2cd34a740824c7877b7dfacba0b71622805d5072655e192937339a1fcef", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 35954, + "estimated_input_tokens": 8989, + "estimated_eval_tokens": 18234 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated bug reports across tokenizers, model-specific generation, trainer state, quantization/offloading, and pipeline utilities. None of the soft pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 43873 is the broadest and most actively discussed report in the set, so it is the best anchor if one issue must be chosen as representative; however, it is not actually a duplicate of the others.", + "canonical_pr_reason": null, + "best_issue_reason": "43873 has the widest scope and strongest ongoing discussion/inbound reference activity, making it the most suitable single issue to represent the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Different bugs: DSA indexer return handling vs DeepSeek-V4 tensor cast/transpose behavior." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "Unrelated code paths: missing ReLU in the DSA indexer vs DeepSeek-V4 tensor ops." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "One is about shard-size defaults in save_pretrained; the other is a Qwen3.5 MoE from_pretrained regression." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "Removal of a config field is unrelated to quantization/offloading failures." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "xLSTM training bugs and Bamba torch.compile SDPA issues are different model/back-end problems." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "Both are GLM-related, but they describe different issues: layernorm epsilon config vs RoPE implementation." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Same model family but not the same defect; one is remote-implementation divergence, the other a specific tensor expression concern." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Saving dequantized models and invalid GenerationConfig are separate failure modes." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "Qwen3_5 image-specific infinite loop vs SAM memory overflow are unrelated." + }, + { + "left": "issue:43986", + "right": "issue:45341", + "accept": false, + "reason": "Video AutoProcessor torchvision dependency error is unrelated to a testing_utils bug." + }, + { + "left": "issue:44246", + "right": "issue:46981", + "accept": false, + "reason": "Slow import and repeated 429 retry behavior are different performance/network issues." + }, + { + "left": "issue:44117", + "right": "issue:46701", + "accept": false, + "reason": "Tokenizer mapping None and image-text-to-text pipeline failure are not the same bug." + }, + { + "left": "issue:43012", + "right": "issue:46821", + "accept": false, + "reason": "A bfloat16 warning during compile is unrelated to TimesFM window_size AttributeError." + }, + { + "left": "issue:30947", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer special-token handling and TrainerState token counting are different subsystems." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken constructor conflict is a different tokenizer bug from mapping lookup returning None." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Missing Qwen3Next projections and missing config field in Qwen3_5MoeVisionConfig are distinct model-definition issues." + }, + { + "left": "issue:43673", + "right": "issue:46123", + "accept": false, + "reason": "Cache handling during chunked prefill is unrelated to MaskGenerationPipeline batch-finalization logic." + }, + { + "left": "issue:30947", + "right": "issue:33453", + "accept": false, + "reason": "Both involve tokenizers, but one is LLaMA3 special-token effects and the other is a generic tokenizer-loading regression." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "Image-patch counting in GLM46V and torch.split return handling in GlmMoeDsaIndexer are different bugs." + }, + { + "left": "issue:46172", + "right": "issue:46710", + "accept": false, + "reason": "A deepseek-v4 tensor-expression issue is not the same as an accuracy regression after upgrading Transformers." + }, + { + "left": "issue:44410", + "right": "issue:45362", + "accept": false, + "reason": "Missing projections in qwen3next and Qwen3.5 chat crashes are different model behaviors." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "trainer.evaluate after train and import failure resolved by uninstalling datasets are unrelated." + }, + { + "left": "issue:46298", + "right": "issue:46752", + "accept": false, + "reason": "Noisy logging and apply_chat_template empty-list crash are not the same issue." + }, + { + "left": "issue:46655", + "right": "issue:46664", + "accept": false, + "reason": "Both are SmolVLM video-processor bugs, but one is padding calculation and the other is num_frames semantics; not the same concrete defect." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "AutoProcessor kwargs passthrough and auto_docstring annotation handling are different code paths." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "Missing torchvision handling and shard-size default behavior are unrelated." + }, + { + "left": "issue:38617", + "right": "issue:41950", + "accept": false, + "reason": "ImportError for layer_type_validation and video-classification pipeline selecting image processors are different regressions." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but they refer to different missing symbols from different dependencies." + }, + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state corruption and Liger Kernel not applying with model_init are separate training 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": 28, + "review_unit_index": 61, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:30947", + "issue:33357", + "issue:36010", + "issue:36683", + "issue:42175", + "issue:42757", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43352", + "issue:43502", + "issue:43526", + "issue:43550", + "issue:43582", + "issue:43756", + "issue:43824", + "issue:43825", + "issue:43827", + "issue:43906", + "issue:43976", + "issue:43992", + "issue:44077", + "issue:44162", + "issue:44368", + "issue:44380", + "issue:44451", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44734", + "issue:44938", + "issue:45479", + "issue:45584", + "issue:45712", + "issue:45753", + "issue:46132", + "issue:46172", + "issue:46178", + "issue:46298", + "issue:46365", + "issue:46421", + "issue:47116", + "issue:47276" + ], + "soft_pairs": [ + "issue:42890|issue:46298", + "issue:44380|issue:46421", + "issue:44451|issue:47116", + "issue:43502|issue:43992", + "issue:44938|issue:46178", + "issue:43208|issue:43526", + "issue:43012|issue:43550", + "issue:43064|issue:47116", + "issue:43066|issue:43906", + "issue:45712|issue:46298", + "issue:42907|issue:45753", + "issue:42907|issue:44655", + "issue:45584|issue:46365", + "issue:43825|issue:44655", + "issue:43012|issue:44368", + "issue:44734|issue:47116", + "issue:43827|issue:47116", + "issue:30947|issue:33357", + "issue:36683|issue:43582", + "issue:44380|issue:45479", + "issue:44077|issue:46365", + "issue:43352|issue:43976", + "issue:43992|issue:44484", + "issue:44380|issue:46132", + "issue:44162|issue:46172", + "issue:42175|issue:46365", + "issue:42757|issue:43824", + "issue:43756|issue:44485", + "issue:43012|issue:47276", + "issue:36010|issue:42757", + "issue:43352|issue:44380", + "issue:43097|issue:46298" + ], + "prepared_review_unit_hash": "54964cc9a6e3552a5ecb73b487ecabefc8a1937ff5b4dac3d8999eafa744e12d", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36581, + "estimated_input_tokens": 9146, + "estimated_eval_tokens": 18548 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose semantic cluster rather than a duplicate set: the items span unrelated import errors, model-specific bugs, pipeline regressions, warnings, and docs/issues. None of the soft pairs look like the same concrete fix or a mergeable PR pair.", + "confidence": 0.9, + "canonical_issue_reason": "No single issue cleanly represents the cluster; the overlap is mostly superficial (shared Transformers/v5 vocabulary) across different code paths and models.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue stands out because the cluster is heterogeneous and does not center on one underlying bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "Flaky test-seed coverage vs a noisy logging message; unrelated failures." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "GPT2 attention scaling ignored vs cache lazy initialization causing recompilation; different subsystems." + }, + { + "left": "issue:44451", + "right": "issue:47116", + "accept": false, + "reason": "Model loading failure vs streaming pipeline regression; no shared concrete bug." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "local_files_only making network requests vs missing embed_tokens.weight during encoder load; distinct issues." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Python 3.14 import/load failure vs bad backend mapping validation; different import paths." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training blockers vs BeitImageProcessorFast label reduction bug; unrelated model/components." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "Compilation warning on bfloat16 vs a Bamba SDPA torch.compile failure; not the same bug." + }, + { + "left": "issue:43064", + "right": "issue:47116", + "accept": false, + "reason": "FSDP2 optimizer-state corruption vs pipeline streaming regression; different code paths." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Tokenizer decoder type regression vs an isolated reproduction of another issue; no shared fix." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Dummy class leakage/check_repo failure vs logging noise; unrelated." + }, + { + "left": "issue:42907", + "right": "issue:45753", + "accept": false, + "reason": "Saving dequantized Ministrals/Devstrals vs Qwen3_5 infinite loop on one image; different bugs." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Dequantized-model save failure vs Pipeline save_pretrained behavior; different objects and failure modes." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation bug vs Trainer import blocked by datasets; unrelated." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Wrong translation-support error message vs inability to save Pipeline objects; both pipeline-related but not the same defect." + }, + { + "left": "issue:43012", + "right": "issue:44368", + "accept": false, + "reason": "bfloat16 compile warning vs tie_word_embeddings warning during LoRA fine-tuning; different warnings and triggers." + }, + { + "left": "issue:44734", + "right": "issue:47116", + "accept": false, + "reason": "KV-cache continuation tensor indexing crash in serve vs streaming pipeline regression; distinct runtime paths." + }, + { + "left": "issue:43827", + "right": "issue:47116", + "accept": false, + "reason": "Docs still referencing removed pipeline APIs vs a runtime streaming regression; docs-only versus code bug." + }, + { + "left": "issue:30947", + "right": "issue:33357", + "accept": false, + "reason": "Special-token flags not taking effect in LLaMA3 vs MacOS bus error in CLIP; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43582", + "accept": false, + "reason": "Gemma3Config missing vocab_size vs Apple Silicon TypeError in allocator warmup; no overlap." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention scaling ignored under SDPA/FlashAttention vs degenerate zero loss in sequence classification; different model behaviors." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "patchtsmixer post_init contract change vs Trainer import conflict with datasets; unrelated." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "NemotronHForCausalLM FlashAttention-2 unsupported vs Python 3.9/3.10 compatibility break; different problems." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing encoder weight on load vs max_shard_size question; not duplicate bugs." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT2 SDPA/FlashAttention scaling bug vs AttentionInterface.register behavior change; both attention-related but separate issues." + }, + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "ESM2 broken broadly vs a deepseek-v4 dtype/transpose bug; different models and code paths." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "Package backend extras issue vs Trainer import blocked by datasets; distinct import/dependency problems." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Missing huggingface_hub symbol vs missing Transformers model class export; different imports." + }, + { + "left": "issue:43756", + "right": "issue:44485", + "accept": false, + "reason": "Smollm3 RoPE layer-count bug vs GLM-5 RoPE implementation discussion; different models and failure modes." + }, + { + "left": "issue:43012", + "right": "issue:47276", + "accept": false, + "reason": "Compile warning on bfloat16 vs misleading synced_gpus warning; only superficially similar as warnings." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Missing GenerationMixin import vs missing is_offline_mode import; separate broken symbols." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder API vs a noisy logging message; 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": 28, + "review_unit_index": 62, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29870", + "issue:29925", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:38617", + "issue:42548", + "issue:42831", + "issue:42890", + "issue:43064", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43525", + "issue:43611", + "issue:43618", + "issue:43761", + "issue:43867", + "issue:43976", + "issue:44077", + "issue:44222", + "issue:44261", + "issue:44263", + "issue:44351", + "issue:44355", + "issue:44380", + "issue:44483", + "issue:44514", + "issue:44625", + "issue:44655", + "issue:44855", + "issue:44938", + "issue:45216", + "issue:45276", + "issue:45478", + "issue:45593", + "issue:45759", + "issue:46077", + "issue:46202", + "issue:46365", + "issue:46512", + "issue:46877", + "issue:47059", + "issue:47276", + "issue:47321", + "issue:47332" + ], + "soft_pairs": [ + "issue:42831|issue:44261", + "issue:43618|issue:43761", + "issue:44483|issue:44514", + "issue:43064|issue:44380", + "issue:44222|issue:45478", + "issue:42890|issue:47059", + "issue:44625|issue:46365", + "issue:45276|issue:46077", + "issue:44655|issue:47321", + "issue:44655|issue:47332", + "issue:38617|issue:43502", + "issue:44625|issue:46077", + "issue:43611|issue:44655", + "issue:43299|issue:43867", + "issue:36683|issue:43475", + "issue:43352|issue:45593", + "issue:44380|issue:47276", + "issue:36683|issue:44077", + "issue:46077|issue:46512", + "issue:44263|issue:45759", + "issue:44355|issue:46877", + "issue:43976|issue:44938", + "issue:29925|issue:33453", + "issue:43867|issue:45216", + "issue:36683|issue:44625", + "issue:44938|issue:47059", + "issue:43618|issue:45216", + "issue:36683|issue:43525", + "issue:44355|issue:44855", + "issue:42548|issue:44351", + "issue:29870|issue:36296", + "issue:43867|issue:46202" + ], + "prepared_review_unit_hash": "e6f57b618664f93565ba75a8faa610c36e13a5bb14bb30679be4c1a2f04a879a", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36504, + "estimated_input_tokens": 9126, + "estimated_eval_tokens": 18508 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: it mixes unrelated model/config regressions, import errors, tokenizer issues, training bugs, and API behavior changes. There is no single duplicate theme strong enough to treat any item as a canonical cluster anchor, and none of the soft pairs look like the same underlying bug/change. No PRs are present in the set.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items span multiple unrelated subsystems and failure modes, so selecting one issue would be arbitrary rather than duplicate-driven.", + "canonical_pr_reason": null, + "best_issue_reason": "If a representative issue is needed, issue 36683 is the strongest standalone exemplar because it is a concrete, reproducible runtime bug report with substantial discussion activity; however, it is not representative enough to serve as a canonical duplicate anchor for this mixed cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42831", + "right": "issue:44261", + "accept": false, + "reason": "Different problems: FineGrainedFP8 accuracy vs MLA q_a_layernorm missing rms_norm_eps. Both are precision-related, but the underlying bugs and fixes are not the same." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": false, + "reason": "Related CLIP area, but one is missing attentions assignment and the other is hidden_states being None in forward. Different outputs, different code paths." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Different request-handling bugs: chat completions API rejection vs batched Qwen2.5-VL chat template crash with padding=False." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "Different training/runtime issues: optimizer state corruption under FSDP2+PEFT vs GPT2 attention scaling ignored under SDPA/FlashAttention." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both mention FP8/MoE, but one is save_pretrained for FP8 moe models and the other is from_pretrained failure for Qwen3.5 MoE. Not the same concrete bug." + }, + { + "left": "issue:42890", + "right": "issue:47059", + "accept": false, + "reason": "One is flaky SAM HQ tests due to missing set_seed; the other is a CLI incompatibility with huggingface_hub 1.22.0. Different failure modes." + }, + { + "left": "issue:44625", + "right": "issue:46365", + "accept": false, + "reason": "Qwen3.5 num_labels propagation bug vs inability to import Trainer until datasets is uninstalled. Unrelated subsystems." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Both are Gemma4-related, but one is resize_token_embeddings not updating embeddings and the other is bidirectional attention still building causal masks. Different code paths." + }, + { + "left": "issue:44655", + "right": "issue:47321", + "accept": false, + "reason": "Pipeline save_pretrained failure vs AMD quark class not updated. No shared underlying bug." + }, + { + "left": "issue:44655", + "right": "issue:47332", + "accept": false, + "reason": "Pipeline save_pretrained failure vs inability to save disk-offloaded Qwen3-VL. Different save/offload behavior and likely different fixes." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "Both involve import/loading behavior, but one is a missing symbol import and the other is local_files_only still making API requests. Not the same issue." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and Gemma4 bidirectional attention are distinct configuration bugs in different model families." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Base_model_prefix loading regression is unrelated to Pipeline save_pretrained behavior." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "Qwen3VL MoE loading regression from v5 vs sorted state_dict load error. Both are loading-related, but not the same code-path problem." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Gemma3Config missing vocab_size is unrelated to SAM 3 Video missing fpn_position_embeddings." + }, + { + "left": "issue:43352", + "right": "issue:45593", + "accept": false, + "reason": "Flash Attention 2 support check for NemotronHForCausalLM vs D-FINE auxiliary losses disabled when denoising is off. Completely different bugs." + }, + { + "left": "issue:44380", + "right": "issue:47276", + "accept": false, + "reason": "GPT2 attention scaling ignored in backends vs misleading synced_gpus warning. One is functional behavior, the other is warning text." + }, + { + "left": "issue:36683", + "right": "issue:44077", + "accept": false, + "reason": "Gemma3Config missing vocab_size and patchtsmixer post_init validation are unrelated model/config issues." + }, + { + "left": "issue:46077", + "right": "issue:46512", + "accept": false, + "reason": "Gemma4 attention mask bug vs from_config(dtype=...) leaving model.config.dtype stale. Different configuration/state propagation bugs." + }, + { + "left": "issue:44263", + "right": "issue:45759", + "accept": false, + "reason": "torch.split return values in GlmMoeDsaIndexer vs AutoModelForCausalLM.from_config not unwrapping text_config for composite Qwen configs. Different model-loading logic." + }, + { + "left": "issue:44355", + "right": "issue:46877", + "accept": false, + "reason": "Compiled Python file import errors vs auto_docstring printing doc-lint diagnostics at import time. Not the same underlying issue." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Python 3.9+ compatibility problem in Transformers 5.1.0 vs Python 3.14 import/load failure. Similar platform topic, but not the same bug." + }, + { + "left": "issue:29925", + "right": "issue:33453", + "accept": false, + "reason": "SigLIP use_fast enforcement vs tokenizer loading regression. The latter is broader and not clearly the same as the former." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "State_dict sorting load error vs Qwen3.5 saved checkpoint regression. Both touch save/load, but the failing mechanism differs." + }, + { + "left": "issue:36683", + "right": "issue:44625", + "accept": false, + "reason": "Gemma3Config vocab_size missing vs Qwen3.5 num_labels propagation. Different configs and different missing fields." + }, + { + "left": "issue:44938", + "right": "issue:47059", + "accept": false, + "reason": "Python 3.14 load failure vs CLI breakage with huggingface_hub 1.22.0. Unrelated compatibility issues." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIP attentions regression vs Qwen3.5 checkpoint save regression. No shared code-path or fix." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3Config missing vocab_size vs Llama4Config missing pad_token_id. Both are missing-config-field errors, but for different models and different fields." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Compiled Python import issues vs torch.jit.script parsing failure on Python 3.13. Different causes and fixes." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "PreTrainedModel import failure vs HybridCache import failure. Same broad import category, but different missing symbols and likely different breakages." + }, + { + "left": "issue:29870", + "right": "issue:36296", + "accept": false, + "reason": "BART generate() decoder_attention_mask bug vs tensor parallel training bug. Completely different model/runtime paths." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "state_dict sorting load error vs Gemma pipeline roundtrip failure on \u2581. Different loading/serialization 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": 28, + "review_unit_index": 63, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28282", + "issue:29870", + "issue:33453", + "issue:36296", + "issue:41628", + "issue:41669", + "issue:42548", + "issue:42890", + "issue:43010", + "issue:43064", + "issue:43502", + "issue:43525", + "issue:43575", + "issue:43618", + "issue:43867", + "issue:43976", + "issue:43992", + "issue:44062", + "issue:44075", + "issue:44077", + "issue:44112", + "issue:44222", + "issue:44230", + "issue:44246", + "issue:44360", + "issue:44393", + "issue:44492", + "issue:44908", + "issue:44964", + "issue:45125", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45440", + "issue:45478", + "issue:45561", + "issue:45850", + "issue:46234", + "issue:46285", + "issue:46291", + "issue:46365", + "issue:46851", + "issue:47000", + "issue:47317", + "issue:47337", + "issue:47455" + ], + "soft_pairs": [ + "issue:44393|issue:47337", + "issue:44908|issue:46851", + "issue:43992|issue:47317", + "issue:43976|issue:47455", + "issue:44393|issue:47000", + "issue:44075|issue:46365", + "issue:43010|issue:45278", + "issue:29870|issue:33453", + "issue:43992|issue:45310", + "issue:36296|issue:43575", + "issue:46291|issue:47455", + "issue:43618|issue:44077", + "issue:43064|issue:44492", + "issue:41669|issue:44246", + "issue:42890|issue:45561", + "issue:28282|issue:33453", + "issue:43867|issue:45357", + "issue:44222|issue:45125", + "issue:43867|issue:45375", + "issue:43010|issue:43976", + "issue:43618|issue:46285", + "issue:46234|issue:47455", + "issue:45125|issue:45478", + "issue:43010|issue:45850", + "issue:45125|issue:45375", + "issue:44360|issue:45440", + "issue:43502|issue:44062", + "issue:42890|issue:44112", + "issue:41628|issue:42548", + "issue:43525|issue:47317", + "issue:44222|issue:44230", + "issue:42890|issue:44964" + ], + "prepared_review_unit_hash": "d47cdb39ed00475c2e3a0b37329b06366a8cf4ed3d385b09ec3d9f61e2a0ab1d", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36265, + "estimated_input_tokens": 9067, + "estimated_eval_tokens": 18390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are related only by broad subsystem similarity, not the same concrete bug. I would deduplicate two pairs only: the Qwen3.5 MoE from_pretrained regression and the import-slowdown reports. No PRs are present.", + "confidence": 0.64, + "canonical_issue_reason": "issue:41669 is the strongest anchor in the set: it describes a concrete import-time regression with an explicit suspected cause, and it cleanly absorbs the matching slowdown report in issue:44246.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:41669 is the most actionable and specific issue overall; it has a clear symptom, a likely root cause, and a plausible duplicate counterpart.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44393", + "right": "issue:47337", + "accept": false, + "reason": "Both mention bbox/output problems, but one is a model output issue and the other is a flaky test; they do not look like the same underlying bug." + }, + { + "left": "issue:44908", + "right": "issue:46851", + "accept": false, + "reason": "Scheduler kwargs handling and Qwen3.5 variable-length position_ids are unrelated bugs." + }, + { + "left": "issue:43992", + "right": "issue:47317", + "accept": false, + "reason": "UMT5Encoder missing weights and a PPFormulaNet training-loss bug are different code paths." + }, + { + "left": "issue:43976", + "right": "issue:47455", + "accept": false, + "reason": "A Python version compatibility issue and a kernel version-check bug are not the same defect." + }, + { + "left": "issue:44393", + "right": "issue:47000", + "accept": false, + "reason": "Both are vision/detection-related, but one is about Qwen3-VL bbox output and the other is RTDetrLoss infeasibility; different failures." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "Optimizer SGD args being ignored and Trainer import failing with datasets are unrelated." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "Cache update no_grad decoration and broad import errors after upgrade are different problems." + }, + { + "left": "issue:29870", + "right": "issue:33453", + "accept": false, + "reason": "BART generate() decoder_attention_mask and tokenizer loading regression are separate issues." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "Missing embed_tokens.weight in UMT5Encoder is not the same as Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor-parallel training state corruption and an OOM while loading a large model are different failure modes." + }, + { + "left": "issue:46291", + "right": "issue:47455", + "accept": false, + "reason": "Importing pipelines with kernels v0.15 and a generic kernel version-check bug are related only at a very high level." + }, + { + "left": "issue:43618", + "right": "issue:44077", + "accept": false, + "reason": "Missing CLIPOutput attentions and patchtsmixer post_init gating are not the same bug." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "FSDP2/PEFT optimizer state corruption and a cache-strategy typo are unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": true, + "reason": "Both describe the same import-time slowdown in transformers; 41669 gives the concrete cause (import * usage), while 44246 is the symptom report." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "A flaky SAM HQ test due to missing set_seed() is unrelated to an xdist race on captured_info.txt." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch for AutoModel and tokenizer loading regression are not the same underlying issue." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "State-dict loading sorted error and incorrect visual encoder keys on save_pretrained are different model-save/load bugs." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "FP8 save_pretrained behavior and a missing tensor-parallel plan for Qwen3_5Moe are separate defects." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "Sorted state_dict loading and missing deepstack_visual_indexes are unrelated." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update decoration and Python version compatibility are unrelated." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "CLIPOutput attentions and CLIPTextModel structure inconsistency are different CLIP-family issues." + }, + { + "left": "issue:46234", + "right": "issue:47455", + "accept": false, + "reason": "Missing import_utils symbol and kernel version-check logic are not the same bug." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": true, + "reason": "These are the same Qwen3.5 MoE from_pretrained error, with near-identical titles and the same model path." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "A cache/Layer no_grad concern and a broad complaint that 5.0 breaks too many models are not the same concrete bug." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Both involve Qwen3.5 MoE vision components, but one is a from_pretrained error and the other is a missing config field." + }, + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "A DSA indexer activation bug and divergence of DeepseekV3MoE from remote implementation are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "Offline-loading API requests and AddedToken special-argument duplication are different code paths." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "A flaky integration test and a CI device-override race are not the same failure." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Missing AutoImageProcessor and missing PreTrainedModel are both import/export regressions, but they are different symbols and likely different fixes." + }, + { + "left": "issue:43525", + "right": "issue:47317", + "accept": false, + "reason": "Llama4Config missing pad_token_id and a PPFormulaNet training-loss bug are unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "FP8 save_pretrained for MoE and fp8 inference support for Qwen3-VL/Qwen3.5 MoE are different requests." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "SAM HQ test flakiness and Phi-4 multimodal model loading are unrelated issues." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The verdicts are grounded in the packet and stay conservative. The two accepted duplicates are well supported by the titles: the Qwen3.5 MoE from_pretrained reports (45125/45478) are effectively the same issue, and the import-slowdown reports (41669/44246) plausibly describe the same regression/symptom pair. The rejected pairs are mostly limited to broad subsystem similarity, which is 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": 28, + "review_unit_index": 64, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:30990", + "issue:31795", + "issue:36296", + "issue:36683", + "issue:38175", + "issue:38617", + "issue:41628", + "issue:42757", + "issue:42898", + "issue:43010", + "issue:43065", + "issue:43232", + "issue:43299", + "issue:43475", + "issue:43502", + "issue:43526", + "issue:43576", + "issue:43723", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:44038", + "issue:44077", + "issue:44230", + "issue:44355", + "issue:44387", + "issue:44485", + "issue:44556", + "issue:44734", + "issue:44749", + "issue:44861", + "issue:44961", + "issue:45071", + "issue:45125", + "issue:45850", + "issue:46095", + "issue:46132", + "issue:46143", + "issue:46144", + "issue:46617", + "issue:46664", + "issue:46726", + "issue:47059", + "issue:47110", + "issue:47317", + "issue:47455" + ], + "soft_pairs": [ + "issue:44387|issue:44556", + "issue:44734|issue:46617", + "issue:41628|issue:47455", + "issue:46143|issue:46144", + "issue:43723|issue:44556", + "issue:43010|issue:43065", + "issue:43824|issue:46617", + "issue:43576|issue:46132", + "issue:42898|issue:46664", + "issue:45071|issue:47317", + "issue:44355|issue:44749", + "issue:43502|issue:44485", + "issue:30990|issue:31795", + "issue:43874|issue:46617", + "issue:43010|issue:43232", + "issue:45850|issue:47110", + "issue:44230|issue:46095", + "issue:43299|issue:44556", + "issue:38175|issue:43502", + "issue:43827|issue:44961", + "issue:46664|issue:46726", + "issue:30333|issue:36296", + "issue:29870|issue:31795", + "issue:44038|issue:45125", + "issue:43576|issue:43824", + "issue:44355|issue:45071", + "issue:43576|issue:45071", + "issue:43502|issue:44961", + "issue:43576|issue:47059", + "issue:36296|issue:43526", + "issue:31795|issue:38617", + "issue:42757|issue:43576", + "issue:36683|issue:47110", + "issue:43475|issue:46617", + "issue:44077|issue:44861" + ], + "prepared_review_unit_hash": "2172d0da17f90db1cf9f302d4734709d7041abc9ea5e48fe9c3cf70b80c1acaf", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 35, + "serialized_chars": 37870, + "estimated_input_tokens": 9468, + "estimated_eval_tokens": 19192 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is a loose collection of Transformers-related regressions, import errors, and docs issues, but the soft pairs are mostly different bugs and should not be merged as duplicates.", + "confidence": 0.79, + "canonical_issue_reason": "issue:45850 is the broadest umbrella report for the general \u201cv5 breaks many models\u201d regression theme.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44556 is the most concrete and reproducible of the v5-loading regression reports, so it is the best representative anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Different failure modes: int4 quantization memory/OOM vs checkpoint reload compatibility after upgrade." + }, + { + "left": "issue:44734", + "right": "issue:46617", + "accept": false, + "reason": "Serve KV-cache indexing crash is unrelated to trust_remote_code cache loading FileNotFoundError." + }, + { + "left": "issue:41628", + "right": "issue:47455", + "accept": false, + "reason": "Importing AutoImageProcessor is unrelated to a kernel version check bug." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Both concern RoFormer, but one is kwargs passthrough and the other is attention-interface integration; not the same bug." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading in v5 is a different code path from reloading old checkpoints after upgrade." + }, + { + "left": "issue:43010", + "right": "issue:43065", + "accept": false, + "reason": "Cache no_grad decoration and Sam3PixelDecoder dummy Conv2d are unrelated changes." + }, + { + "left": "issue:43824", + "right": "issue:46617", + "accept": false, + "reason": "Missing Qwen2_5_VL export and loading a trust_remote_code model from cache are different import/load problems." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "CLI env command breakage is unrelated to AttentionInterface.register behavior changes." + }, + { + "left": "issue:42898", + "right": "issue:46664", + "accept": false, + "reason": "Tokenization-space cleanup behavior and video num_frames semantics are unrelated." + }, + { + "left": "issue:45071", + "right": "issue:47317", + "accept": false, + "reason": "PretrainedConfig type checking and PPFormulaNet training-loss bug are different issues." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Compiled Python file errors are unrelated to preprocessing slowdown after version upgrade." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "local_files_only network requests and GLM-5 RoPE implementation are unrelated." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Sentence-Transformer loading stall is not the same as confusing forward docs." + }, + { + "left": "issue:43874", + "right": "issue:46617", + "accept": false, + "reason": "Missing image-patch helper in GLM46V is unrelated to trust_remote_code cache loading." + }, + { + "left": "issue:43010", + "right": "issue:43232", + "accept": false, + "reason": "Both involve generation internals, but they describe different failures and code paths." + }, + { + "left": "issue:45850", + "right": "issue:47110", + "accept": false, + "reason": "Broad v5 regressions vs a special_tokens dict/list attribute error; not the same underlying bug." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support request and weight-conversion question are different feature topics." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Qwen3VL-Moe loading breakage and old-checkpoint reload regression overlap broadly but are distinct reports." + }, + { + "left": "issue:38175", + "right": "issue:43502", + "accept": false, + "reason": "Zero probabilities in SigLIP2 is unrelated to unexpected API calls with local_files_only." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs references to removed pipeline APIs are unrelated to the 'racoon' issue." + }, + { + "left": "issue:46664", + "right": "issue:46726", + "accept": false, + "reason": "Video num_frames semantics and missing vision backend handling are different processor issues." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLFlow reporting on failed jobs is unrelated to tensor parallel training." + }, + { + "left": "issue:29870", + "right": "issue:31795", + "accept": false, + "reason": "BART generate() runtime error is unrelated to forward-argument documentation confusion." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Both are Qwen3-Moe related, but one is a broad v5 bug and the other is a missing tensor-parallel plan." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "CLI env command breakage and Qwen VL class import failure are unrelated." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file execution errors are unrelated to PretrainedConfig type checking." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "Network access despite local_files_only and 'racoon' are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:47059", + "accept": false, + "reason": "transformers env CLI breakage and huggingface_hub completion attribute error are different dependency issues." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor parallel training bug and BeitImageProcessorFast reduce_labels bug are unrelated." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Docs confusion and layer_type_validation import error do not share the same bug." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "Missing is_offline_mode import from huggingface_hub is unrelated to the env CLI command issue." + }, + { + "left": "issue:36683", + "right": "issue:47110", + "accept": false, + "reason": "Gemma3Config missing vocab_size and special_tokens keys error are different attribute problems." + }, + { + "left": "issue:43475", + "right": "issue:46617", + "accept": false, + "reason": "SAM 3 Video missing fpn_position_embeddings is unrelated to trust_remote_code cache loading." + }, + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Optional post_init acceptance and tied-weights list handling are separate configuration/utility 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": 28, + "review_unit_index": 65, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:41669", + "issue:42371", + "issue:43298", + "issue:43352", + "issue:43408", + "issue:43525", + "issue:43572", + "issue:43673", + "issue:43867", + "issue:43976", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44393", + "issue:44466", + "issue:44485", + "issue:44492", + "issue:44556", + "issue:44623", + "issue:44655", + "issue:44861", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45230", + "issue:45362", + "issue:45375", + "issue:45381", + "issue:45468", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46172", + "issue:46209", + "issue:46218", + "issue:46328", + "issue:46617", + "issue:46664", + "issue:47059", + "issue:47110", + "issue:47455" + ], + "soft_pairs": [ + "issue:45362|issue:45850", + "issue:44351|issue:46617", + "issue:44623|issue:45230", + "issue:44315|issue:44360", + "issue:45375|issue:47110", + "issue:43673|issue:44492", + "issue:45381|issue:46664", + "issue:44273|issue:45542", + "issue:44485|issue:44961", + "issue:43525|issue:47110", + "issue:43352|issue:44961", + "issue:44315|issue:46209", + "issue:44961|issue:45468", + "issue:44273|issue:45003", + "issue:44466|issue:44556", + "issue:43867|issue:47110", + "issue:41669|issue:43673", + "issue:46018|issue:46172", + "issue:43572|issue:45850", + "issue:43525|issue:45850", + "issue:44655|issue:44861", + "issue:44393|issue:45753", + "issue:44485|issue:46664", + "issue:44393|issue:45071", + "issue:44556|issue:45092", + "issue:42371|issue:44492", + "issue:44295|issue:45230", + "issue:44297|issue:45230", + "issue:44246|issue:44273", + "issue:45850|issue:47059", + "issue:43408|issue:44315", + "issue:41669|issue:47455", + "issue:43298|issue:46095", + "issue:44492|issue:46328", + "issue:43976|issue:44393", + "issue:44492|issue:45850", + "issue:41669|issue:46218" + ], + "prepared_review_unit_hash": "12b1c11621379257835602d79911be6c52abbbc52ffe35073f5833e74ba58947", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 37, + "serialized_chars": 38047, + "estimated_input_tokens": 9512, + "estimated_eval_tokens": 19280 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: many items are different model- or feature-specific bugs that only share a loose Transformers v5/loading theme. The only clear duplicate-like edge is the lazy-loading/import-time slowdown pair.", + "confidence": 0.43, + "canonical_issue_reason": "issue 44273 is the most concrete root-cause issue here: it directly describes lazy loading being broken, which plausibly subsumes the slower-import symptom in 44246. The other issues are mostly separate bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "issue 44273 is the best representative issue for the only strong duplicate subcluster, and it is specific enough to anchor a fix without being as generic as the broad v5 regression reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Generic 'v5 breaks models' complaint versus a specific Qwen crash; same theme, not the same bug." + }, + { + "left": "issue:44351", + "right": "issue:46617", + "accept": false, + "reason": "Both are load/import failures, but one is missing HybridCache import and the other is cache FileNotFoundError; different code paths." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "A concrete save_pretrained files bug versus a placeholder bug report with no matching failure mode." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Unrelated issues in different subsystems; no shared concrete bug." + }, + { + "left": "issue:45375", + "right": "issue:47110", + "accept": false, + "reason": "Different problems: strict field dropping in a config versus special_tokens being a list." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache regression versus a cache-strategy typo; not the same underlying defect." + }, + { + "left": "issue:45381", + "right": "issue:46664", + "accept": false, + "reason": "Both concern video/vision processing, but they are different model-specific bugs with different symptoms." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy-loading/import performance issue versus a TensorFlow backend installation bug." + }, + { + "left": "issue:44485", + "right": "issue:44961", + "accept": false, + "reason": "No meaningful overlap beyond being issue titles; unrelated bugs." + }, + { + "left": "issue:43525", + "right": "issue:47110", + "accept": false, + "reason": "Missing pad_token_id in a config is unrelated to a special_tokens dict/list error." + }, + { + "left": "issue:43352", + "right": "issue:44961", + "accept": false, + "reason": "Flash-Attention support warning versus an unrelated issue title." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger kernel application bug versus FP8 inference on B200; different layers and failure modes." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "No common bug; one is a placeholder title, the other is Gemma audio positional encoding." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Both touch import/loading behavior, but one is lazy loading performance and the other is unsafe sys.modules access." + }, + { + "left": "issue:44466", + "right": "issue:44556", + "accept": false, + "reason": "Related to serialization/loading generally, but one is tied-weight serialization inconsistency and the other is checkpoint reload incompatibility." + }, + { + "left": "issue:43867", + "right": "issue:47110", + "accept": false, + "reason": "State_dict sorting load error versus special_tokens keys error; unrelated." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import-star slowdown and chunked_prefill cache issues are both regressions, but not the same concrete bug." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "Both are DeepSeek-V4 related, but one is gating logic and the other is a tensor transpose/cast issue." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "A specific StableLm config regression is not the same as a broad 'v5 breaks many models' report." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "Specific config field regression versus a generic platform-wide complaint." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Both involve save_pretrained internals, but one is pipeline saving and the other is tied-weight-key handling." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both are Qwen vision issues, but the symptoms and likely fixes differ." + }, + { + "left": "issue:44485", + "right": "issue:46664", + "accept": false, + "reason": "Completely different model families and bugs." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Model-specific bounding box bug versus a global PretrainedConfig type-checking regression." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are checkpoint-loading regressions after v5, but the titles point to different failure modes and models; not clearly one bug." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API guidance is unrelated to cache strategy typos." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "Specific position_ids buffer bug versus a generic bug report." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer save_pretrained mismatch versus a generic bug report." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": true, + "reason": "These describe the same underlying lazy-loading regression: broken lazy loading manifests as intermittent slow imports." + }, + { + "left": "issue:45850", + "right": "issue:47059", + "accept": false, + "reason": "Broad v5 regression complaint versus a specific CLI compatibility failure with huggingface_hub 1.22.0." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "Different model warning versus a training-time Liger kernel issue." + }, + { + "left": "issue:41669", + "right": "issue:47455", + "accept": false, + "reason": "Import-star performance regression versus an unrelated kernel-version check issue." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching support for audio models is unrelated to a DeepSeek-V4 weight-conversion question." + }, + { + "left": "issue:44492", + "right": "issue:46328", + "accept": false, + "reason": "Cache-strategy typo versus SequenceFeatureExtractor padding performance; different code paths." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility regression versus a Qwen vision output bug." + }, + { + "left": "issue:44492", + "right": "issue:45850", + "accept": false, + "reason": "A typo in cache strategies is not the same as the broad 'v5 breaks too many models' complaint." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import-time slowdown from import * is unrelated to the Qwen3_5 import failure." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats the cluster as mostly heterogeneous, identifies the 44246\u219444273 lazy-loading/import slowdown pair as the only clear duplicate-like edge, and does not overclaim the weaker similarities. The soft-edge rejections are generally well supported by the issue titles." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 66, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:28348", + "issue:31795", + "issue:33453", + "issue:37428", + "issue:41669", + "issue:42371", + "issue:42890", + "issue:42913", + "issue:43066", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43576", + "issue:43606", + "issue:43618", + "issue:43704", + "issue:43827", + "issue:43976", + "issue:44230", + "issue:44261", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44704", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45071", + "issue:45230", + "issue:45250", + "issue:45290", + "issue:45310", + "issue:45341", + "issue:45468", + "issue:45656", + "issue:45850", + "issue:46218", + "issue:46655", + "issue:46894", + "issue:47059", + "issue:47110", + "issue:47176", + "issue:47177", + "issue:47455" + ], + "soft_pairs": [ + "issue:44704|issue:44908", + "issue:42890|issue:44908", + "issue:41669|issue:45341", + "issue:44908|issue:45071", + "issue:43976|issue:46218", + "issue:43827|issue:44908", + "issue:43576|issue:47176", + "issue:45230|issue:45656", + "issue:42913|issue:43066", + "issue:43976|issue:47176", + "issue:31795|issue:33453", + "issue:44273|issue:47455", + "issue:44315|issue:47455", + "issue:43519|issue:46655", + "issue:45468|issue:46894", + "issue:44861|issue:47110", + "issue:43976|issue:47177", + "issue:44485|issue:47176", + "issue:31795|issue:37428", + "issue:44360|issue:45468", + "issue:43519|issue:44485", + "issue:43576|issue:44273", + "issue:44485|issue:47177", + "issue:44861|issue:47059", + "issue:43519|issue:44908", + "issue:41669|issue:43704", + "issue:43606|issue:45468", + "issue:43576|issue:46218", + "issue:28348|issue:45250", + "issue:42371|issue:43704", + "issue:44485|issue:45468", + "issue:43704|issue:44485", + "issue:43352|issue:47177", + "issue:43352|issue:47176", + "issue:45850|issue:47177", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:44261|issue:45468", + "issue:45850|issue:46218", + "issue:43704|issue:45341", + "issue:43352|issue:46218", + "issue:44351|issue:46218", + "issue:44230|issue:44623", + "issue:44861|issue:45290", + "issue:28348|issue:43352", + "issue:44230|issue:45310", + "issue:44908|issue:44998", + "issue:44998|issue:47177", + "issue:44998|issue:47176", + "issue:44492|issue:44998", + "issue:43618|issue:44998" + ], + "prepared_review_unit_hash": "c7d2bb75b389c6f861fe9398f65462e2d10d44102952f1907005b715c5d67c27", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 51, + "serialized_chars": 42040, + "estimated_input_tokens": 10510, + "estimated_eval_tokens": 21276 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous: most items are unrelated bug reports across tokenization, loading, docs, security, processor, and model-specific regressions. The soft pairs are mostly superficial topical matches, not true duplicate artifacts.", + "confidence": 0.93, + "canonical_issue_reason": "No true canonical issue: the set is too mixed, and the soft similarities look like keyword/topic overlap rather than the same underlying bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43502 is the clearest standalone bug report with a concrete, reproducible loading regression, but it is still not representative of the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "Different code paths: cached_file kwargs passthrough vs scheduler kwargs handling." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Training/test flakiness from missing set_seed is unrelated to scheduler behavior." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "Import-star cleanup and a testing_utils bug are unrelated changes." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs bug is not the same as PretrainedConfig type-checking regression." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python version compatibility and missing Qwen3_5 import are different failures." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Docs still mentioning pipeline() is unrelated to scheduler kwargs handling." + }, + { + "left": "issue:43576", + "right": "issue:47176", + "accept": false, + "reason": "CLI env command breakage and shell-operator bypass are different issues." + }, + { + "left": "issue:45230", + "right": "issue:45656", + "accept": false, + "reason": "Generic bug report versus double optimizer step; no shared code path." + }, + { + "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." + }, + { + "left": "issue:43976", + "right": "issue:47177", + "accept": false, + "reason": "Python support mismatch is unrelated to tar/zip file-write security." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Documentation confusion about forward args is not a tokenizer loading regression." + }, + { + "left": "issue:44273", + "right": "issue:47455", + "accept": false, + "reason": "Lazy loading failure and kernel version checking are unrelated." + }, + { + "left": "issue:44315", + "right": "issue:47455", + "accept": false, + "reason": "Liger Kernel application under model_init is unrelated to version validation." + }, + { + "left": "issue:43519", + "right": "issue:46655", + "accept": false, + "reason": "Both are processor bugs, but timestamp math and video padding are different logic paths." + }, + { + "left": "issue:45468", + "right": "issue:46894", + "accept": false, + "reason": "Gemma audio positional encoding and ASR loss shifting are unrelated model bugs." + }, + { + "left": "issue:44861", + "right": "issue:47110", + "accept": false, + "reason": "Weight-key list crash is not the same as special_tokens list handling." + }, + { + "left": "issue:43976", + "right": "issue:47177", + "accept": false, + "reason": "Python compatibility issue is unrelated to shell/file extraction exploits." + }, + { + "left": "issue:44485", + "right": "issue:47176", + "accept": false, + "reason": "RoPE implementation bug and command-substitution allowlist bypass are unrelated." + }, + { + "left": "issue:31795", + "right": "issue:37428", + "accept": false, + "reason": "Forward-argument docs confusion is unrelated to missing flash-support import symbol." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA ReLU omission is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Processor timestamp math and GLM-5 RoPE are different subsystems." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "CLI env command failure does not match lazy-loading behavior." + }, + { + "left": "issue:44485", + "right": "issue:47177", + "accept": false, + "reason": "Model RoPE math and shell injection/file-write security are unrelated." + }, + { + "left": "issue:44861", + "right": "issue:47059", + "accept": false, + "reason": "Weight-key attribute crash is unrelated to Hugging Face CLI completion breakage." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Processor timestamp calculation is unrelated to scheduler kwargs." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star removal is unrelated to a VRAM leak in Qwen3ForCausalLM." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch and Gemma audio positional encoding are unrelated." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Broken env command and missing model import are different failures." + }, + { + "left": "issue:28348", + "right": "issue:45250", + "accept": false, + "reason": "Both mention Flash Attention 2.0, but one is GPT2-specific support and the other is too generic to treat as the same bug/change." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 API guidance and VRAM leakage are unrelated." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Different model math issues with no shared code path." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in multithreaded dataloaders is unrelated to RoPE implementation." + }, + { + "left": "issue:43352", + "right": "issue:47177", + "accept": false, + "reason": "Flash-Attention support limitation and archive extraction RCE are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:47176", + "accept": false, + "reason": "Model Flash Attention support and shell-operator bypass are unrelated." + }, + { + "left": "issue:45850", + "right": "issue:47177", + "accept": false, + "reason": "Broad version-breakage complaint and file-write/RCE issue are not duplicates." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "Local_files_only network requests and processor timestamp math are different bugs." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Processor timestamp bug is unrelated to missing processor.save_pretrained files." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "RMS norm eps precision bug is unrelated to audio positional encoding." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "General v5 breakage is too broad to merge with a specific missing import error." + }, + { + "left": "issue:43704", + "right": "issue:45341", + "accept": false, + "reason": "VRAM leak in training threads and a testing_utils bug are unrelated." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention support for NemotronH is unrelated to missing Qwen3_5 import." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "HybridCache import failure and Qwen3_5 import failure are different symbols and code paths." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 model support and missing processor save files are unrelated." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Weight-key crash is unrelated to chat template tool-call handling." + }, + { + "left": "issue:28348", + "right": "issue:43352", + "accept": false, + "reason": "Both are Flash Attention 2.0 related, but one is GPT2 support and the other is Nemotron-specific lack of support." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 support request and Qwen3.5 MoE from_pretrained error are not the same bug." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug and an unrelated placeholder issue are not duplicates." + }, + { + "left": "issue:44998", + "right": "issue:47177", + "accept": false, + "reason": "Placeholder issue is unrelated to archive extraction security." + }, + { + "left": "issue:44998", + "right": "issue:47176", + "accept": false, + "reason": "Placeholder issue is unrelated to command-substitution bypass." + }, + { + "left": "issue:44492", + "right": "issue:44998", + "accept": false, + "reason": "Cache-strategy typo and unrelated placeholder issue are not duplicates." + }, + { + "left": "issue:43618", + "right": "issue:44998", + "accept": false, + "reason": "CLIP attentions regression is unrelated to the placeholder issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 28, + "review_unit_index": 67, + "review_unit_count": 67, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-544", + "nodes": [ + "issue:29870", + "issue:30333", + "issue:31515", + "issue:31795", + "issue:36296", + "issue:37934", + "issue:38617", + "issue:42936", + "issue:43240", + "issue:43298", + "issue:43668", + "issue:43717", + "issue:43813", + "issue:44031", + "issue:44230", + "issue:46018", + "issue:46144" + ], + "soft_pairs": [ + "issue:31515|issue:38617", + "issue:43298|issue:44230", + "issue:29870|issue:31515", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:31515|issue:31795", + "issue:43717|issue:46144", + "issue:42936|issue:46018", + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43668|issue:43813", + "issue:43240|issue:43668", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "0cf3db67b7beda680a3d6b2e63ebad817ee96ff52d4df6e1df30801d06b9ed41", + "component_budget": { + "node_count": 544, + "item_count": 544, + "soft_pair_count": 2019, + "serialized_chars": 1044631, + "estimated_input_tokens": 261158, + "estimated_eval_tokens": 522572 + }, + "budget": { + "node_count": 17, + "item_count": 17, + "soft_pair_count": 18, + "serialized_chars": 14985, + "estimated_input_tokens": 3747, + "estimated_eval_tokens": 7750 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All proposed soft edges look like false positives: the issues span unrelated subsystems and symptoms (loading, training, docs, config typing, model behavior, tokenizer warnings, etc.). None of the pairs appear to describe the same underlying bug or change.", + "confidence": 0.98, + "canonical_issue_reason": "No clear canonical issue stands out because the cluster is heterogeneous; issue:31515 is only the most connected by soft similarity, not a semantic duplicate target for the others.", + "canonical_pr_reason": null, + "best_issue_reason": "No global best issue for consolidation: none of the items are true duplicates, so selecting any single issue would misrepresent the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Different problems: checkpoint loading performance vs an import error in configuration utilities." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Unrelated topics: continuous batching audio support vs fp8 inference support for Qwen VL/MoE models." + }, + { + "left": "issue:29870", + "right": "issue:31515", + "accept": false, + "reason": "BART generate() UnboundLocalError is unrelated to slow from_pretrained checkpoint loading." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLFlow reporting on failed training jobs is a separate tracking/reporting issue, not checkpoint loading." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint loading speed and tensor-parallel training bugs are different code paths and failures." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Documentation confusion about forward arguments is not the same as slow checkpoint loading." + }, + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Mamba initialization behavior and RoFormer attention-interface usage are different model-specific implementation issues." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2Former ignore_value handling and DeepSeek-V4 shared-expert gating are unrelated model bugs." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "A config type-hint typo is unrelated to tokenizer regex warning behavior." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Missing ignore_value usage in Mask2Former is not the same as kwargs propagation to a loss function." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss kwargs plumbing and RoFormer attention interface usage affect different parts of the stack." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2Former loss-value handling is unrelated to tokenizer regex warnings." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "These concern different models and bugs: Mask2Former ignore_value vs RoFormer attention API usage." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2Former loss handling and a quantization-operation typo are different issues." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "Incorrect type hint in ModernBERTConfig is unrelated to a typo in quantization conversion code." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Passing kwargs to loss calculation is unrelated to a config annotation problem." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "A question about Llama4TextL2Norm behavior is not the same as a ModernBERT config type-hint bug." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2Former ignore_value usage and ModernBERT norm_eps typing are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 28, + "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": "6bf874f3c6793fe26c857b2ac4fbc84dcfc4821dbf3d416c871fc07a189840c0", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2875, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2875, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The cluster is not a duplicate set: it mixes three unrelated feature requests in tokenization, learning-rate scheduling, and Wav2Vec2 alignment. No pair looks like the same underlying issue.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:11307 has the strongest discussion activity, but it is still unrelated to the other two issues, so there is no true canonical issue for this cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "10519 is about truncating from the beginning in tokenization; 11307 is about getting word start/end offsets in Wav2Vec2. Different components and different problems." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "10519 requests a truncation direction option, while 11005 asks for ReduceLROnPlateau-like scheduler behavior. No shared code path or underlying bug." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "11005 concerns training scheduler functionality; 11307 concerns speech-model time offsets. These are unrelated feature requests." + } + ] + }, + "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": 28, + "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": "ee4e442dc388044b6a830d55d6b5144b359990322fd242076dde0bdc0ed2ffd8", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues do not appear to describe the same underlying bug: one is about Tapas tokenization behavior, one about an authentication/error-message problem when downloading private models, and one about loading bert-base-uncased. The soft links look like broad topical similarity only.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization mismatch and generic model loading failure are different code-path problems; no clear shared root cause." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "Private-model login/download error messaging is distinct from a general inability to load bert-base-uncased." + } + ] + }, + "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": 28, + "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-19631-10", + "nodes": [ + "issue:19631", + "issue:19865", + "issue:20334", + "issue:20362", + "issue:22352", + "issue:22372", + "issue:23050", + "issue:25725", + "issue:25839", + "issue:28306" + ], + "soft_pairs": [ + "issue:19631|issue:22372", + "issue:19631|issue:20362", + "issue:19865|issue:25725", + "issue:19865|issue:22372", + "issue:20334|issue:22352", + "issue:20334|issue:20362", + "issue:20334|issue:25725", + "issue:20362|issue:25839", + "issue:20362|issue:25725", + "issue:22352|issue:25725", + "issue:22372|issue:25725", + "issue:23050|issue:28306", + "issue:23050|issue:25725" + ], + "prepared_review_unit_hash": "41ef4761d7e29dfafe647657336b00a125e19dff5d21d292d2537080f973b768", + "component_budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 13, + "serialized_chars": 9966, + "estimated_input_tokens": 2492, + "estimated_eval_tokens": 5240 + }, + "budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 13, + "serialized_chars": 9966, + "estimated_input_tokens": 2492, + "estimated_eval_tokens": 5240 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues are all distinct feature requests for different models or model-related tasks. The soft similarities come from being in the same broad \u2018add a model\u2019 area, but there is no evidence they describe the same underlying change or bug.", + "confidence": 0.96, + "canonical_issue_reason": "No true canonical duplicate: the cluster is heterogeneous, covering unrelated model additions (EDSR/MDSR, VATT, FlexiBERT, Restormer, LLaMA-MoE, etc.). If forced to pick the broadest issue, issue:25725 (Meta-Transformer) is the most generic, but it does not subsume the others.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:25725 is the broadest/most generic model-request title in the set, but it is still not a duplicate representative for the rest; the items are separate requests for different architectures and tasks.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:19631", + "right": "issue:22372", + "accept": false, + "reason": "EDSR/MDSR super-resolution and Restormer are different model families and different requested additions." + }, + { + "left": "issue:19631", + "right": "issue:20362", + "accept": false, + "reason": "EDSR/MDSR and FlexiBERT are unrelated model requests with no shared code-path or concrete duplicate target." + }, + { + "left": "issue:19865", + "right": "issue:25725", + "accept": false, + "reason": "VATT and Meta-Transformer are different research models; similarity is only that both are model-addition requests." + }, + { + "left": "issue:19865", + "right": "issue:22372", + "accept": false, + "reason": "VATT (audio/vision transformer) and Restormer (image restoration) are distinct changes." + }, + { + "left": "issue:20334", + "right": "issue:22352", + "accept": false, + "reason": "longformer_content and Whisper XVector are unrelated tasks and model integrations." + }, + { + "left": "issue:20334", + "right": "issue:20362", + "accept": false, + "reason": "longformer_content and FlexiBERT are different model requests, not the same underlying change." + }, + { + "left": "issue:20334", + "right": "issue:25725", + "accept": false, + "reason": "longformer_content and Meta-Transformer refer to different models and use cases." + }, + { + "left": "issue:20362", + "right": "issue:25839", + "accept": false, + "reason": "FlexiBERT and D_Nikud are separate model additions for different NLP tasks." + }, + { + "left": "issue:20362", + "right": "issue:25725", + "accept": false, + "reason": "FlexiBERT and Meta-Transformer are different architectures and request contexts." + }, + { + "left": "issue:22352", + "right": "issue:25725", + "accept": false, + "reason": "Whisper XVector finetuning and Meta-Transformer are unrelated; they do not look mergeable into one fix." + }, + { + "left": "issue:22372", + "right": "issue:25725", + "accept": false, + "reason": "Restormer and Meta-Transformer are different models serving different purposes." + }, + { + "left": "issue:23050", + "right": "issue:28306", + "accept": false, + "reason": "TTS advanced and LLaMA-MoE are separate model requests in different domains." + }, + { + "left": "issue:23050", + "right": "issue:25725", + "accept": false, + "reason": "Advanced TTS and Meta-Transformer are not the same underlying feature or code path." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 5, + "component_count": 28, + "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-26350-16", + "nodes": [ + "issue:26350", + "issue:26794", + "issue:27088", + "issue:28936", + "issue:30191", + "issue:30240", + "issue:33260", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370", + "issue:46222", + "issue:46226", + "issue:46931", + "issue:47299", + "issue:47402" + ], + "soft_pairs": [ + "issue:26350|issue:33260", + "issue:26794|issue:44370", + "issue:26794|issue:44365", + "issue:26794|issue:47402", + "issue:26794|issue:43834", + "issue:26794|issue:47299", + "issue:26794|issue:27088", + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:47402", + "issue:27088|issue:43834", + "issue:28936|issue:43834", + "issue:27088|issue:28936", + "issue:28936|issue:44365", + "issue:28936|issue:30240", + "issue:28936|issue:38435", + "issue:27088|issue:30191", + "issue:26794|issue:30191", + "issue:30191|issue:30240", + "issue:28936|issue:30191", + "issue:26794|issue:30240", + "issue:30240|issue:44370", + "issue:30240|issue:44365", + "issue:30240|issue:47402", + "issue:30240|issue:43834", + "issue:38435|issue:44370", + "issue:38435|issue:44365", + "issue:38435|issue:47402", + "issue:38435|issue:43834", + "issue:38435|issue:47299", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:43834|issue:47402", + "issue:43834|issue:47299", + "issue:43834|issue:46226", + "issue:44365|issue:44370", + "issue:44365|issue:47402", + "issue:44365|issue:47299", + "issue:44365|issue:46226", + "issue:44370|issue:47402", + "issue:44370|issue:47299", + "issue:44370|issue:46226", + "issue:46222|issue:47299", + "issue:44370|issue:46222", + "issue:44365|issue:46222", + "issue:46222|issue:47402", + "issue:43834|issue:46222", + "issue:46226|issue:47402", + "issue:46226|issue:46931", + "issue:44370|issue:46931", + "issue:44365|issue:46931", + "issue:43834|issue:46931", + "issue:46931|issue:47402", + "issue:47299|issue:47402", + "issue:46226|issue:47299" + ], + "prepared_review_unit_hash": "3a0a7a40003b31b451a4c64578697f1c10bcb8ae7ea48dfde5f09d44a6129d5b", + "component_budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 55, + "serialized_chars": 29306, + "estimated_input_tokens": 7327, + "estimated_eval_tokens": 14910 + }, + "budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 55, + "serialized_chars": 29306, + "estimated_input_tokens": 7327, + "estimated_eval_tokens": 14910 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a set of distinct i18n translation-request issues with highly similar template wording, plus two unrelated feature requests (Flash Attention 2 and GGUF support). None of the soft edges look like true duplicates of the same underlying change.", + "confidence": 0.88, + "canonical_issue_reason": "issue:43834 is the cleanest generic i18n template and the best representative of the repeated translation-request family, but the neighboring issues still appear to be separate language-specific requests rather than duplicates.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43834 is the most central and most broadly similar issue in the translation subgroup, so it is the best single issue to anchor the cluster even though it should not absorb the others as duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Different feature areas: Flash Attention 2 support vs GGUF support. Same broad contribution style, but not the same underlying change." + }, + { + "left": "issue:26794", + "right": "issue:44370", + "accept": false, + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." + }, + { + "left": "issue:26794", + "right": "issue:44365", + "accept": false, + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." + }, + { + "left": "issue:26794", + "right": "issue:47402", + "accept": false, + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." + }, + { + "left": "issue:26794", + "right": "issue:43834", + "accept": false, + "reason": "Both are i18n translation requests, but likely for different languages; similar template wording is not enough to treat them as the same issue." + }, + { + "left": "issue:26794", + "right": "issue:47299", + "accept": false, + "reason": "The titles do not indicate the same language or request; no evidence these are the same translation task." + }, + { + "left": "issue:26794", + "right": "issue:27088", + "accept": false, + "reason": "Different target languages are implied; these are separate translation requests, not a duplicate." + }, + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." + }, + { + "left": "issue:27088", + "right": "issue:47402", + "accept": false, + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Both look like translation requests, but they appear to be distinct language-specific issues rather than one underlying task." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish translation vs a generic i18n template issue; same subsystem, but not the same language request." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Turkish and Spanish translation requests are separate language tasks." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Different translation requests; no evidence they refer to the same language or work item." + }, + { + "left": "issue:28936", + "right": "issue:30240", + "accept": false, + "reason": "Spanish and Polish translation requests are separate language tasks." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Spanish and Romanian translation requests are separate language tasks." + }, + { + "left": "issue:27088", + "right": "issue:30191", + "accept": false, + "reason": "Turkish and Persian translation requests are separate language tasks." + }, + { + "left": "issue:26794", + "right": "issue:30191", + "accept": false, + "reason": "Kannada and Persian translation requests are separate language tasks." + }, + { + "left": "issue:30191", + "right": "issue:30240", + "accept": false, + "reason": "Persian and Polish translation requests are separate language tasks." + }, + { + "left": "issue:28936", + "right": "issue:30191", + "accept": false, + "reason": "Spanish and Persian translation requests are separate language tasks." + }, + { + "left": "issue:26794", + "right": "issue:30240", + "accept": false, + "reason": "Kannada and Polish translation requests are separate language tasks." + }, + { + "left": "issue:30240", + "right": "issue:44370", + "accept": false, + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:30240", + "right": "issue:44365", + "accept": false, + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:30240", + "right": "issue:47402", + "accept": false, + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:30240", + "right": "issue:43834", + "accept": false, + "reason": "Polish translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:38435", + "right": "issue:47402", + "accept": false, + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian translation request versus a different generic i18n template issue; not the same underlying request." + }, + { + "left": "issue:38435", + "right": "issue:47299", + "accept": false, + "reason": "Different translation request content; no sign these are the same underlying issue." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." + }, + { + "left": "issue:43834", + "right": "issue:47402", + "accept": false, + "reason": "They share the same template language, but there is no evidence they are the same language-specific request." + }, + { + "left": "issue:43834", + "right": "issue:47299", + "accept": false, + "reason": "Template similarity alone is not enough; the underlying translation request is not shown to be the same." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "Different issue text and no evidence of the same translation request." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." + }, + { + "left": "issue:44365", + "right": "issue:47402", + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." + }, + { + "left": "issue:44365", + "right": "issue:47299", + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Different issue text and no evidence they are the same translation task." + }, + { + "left": "issue:44370", + "right": "issue:47402", + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." + }, + { + "left": "issue:44370", + "right": "issue:47299", + "accept": false, + "reason": "Identical template wording does not prove the same language request; treat as separate issues." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "Different issue text and no evidence they are the same translation task." + }, + { + "left": "issue:46222", + "right": "issue:47299", + "accept": false, + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they represent the same request." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:46222", + "right": "issue:47402", + "accept": false, + "reason": "No evidence these are the same underlying translation request." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:46226", + "right": "issue:47402", + "accept": false, + "reason": "No evidence these are the same underlying translation request." + }, + { + "left": "issue:46226", + "right": "issue:46931", + "accept": false, + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they describe the same request." + }, + { + "left": "issue:44370", + "right": "issue:46931", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:44365", + "right": "issue:46931", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:43834", + "right": "issue:46931", + "accept": false, + "reason": "Generic translation template issue versus unrelated nonstandard text; not enough to call it a duplicate." + }, + { + "left": "issue:46931", + "right": "issue:47402", + "accept": false, + "reason": "Different titles and no evidence of the same underlying translation request." + }, + { + "left": "issue:47299", + "right": "issue:47402", + "accept": false, + "reason": "Different titles and no evidence of the same underlying translation request." + }, + { + "left": "issue:46226", + "right": "issue:47299", + "accept": false, + "reason": "Both are nonstandard-language titles, but there is no reliable evidence they represent the same request." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 6, + "component_count": 28, + "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-30396-2", + "nodes": [ + "issue:30396", + "issue:31103" + ], + "soft_pairs": [ + "issue:30396|issue:31103" + ], + "prepared_review_unit_hash": "37855664083aa72d66d91c6fdc50d022f0c490c43cfb54cb156b83e23beb38ac", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1539, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1539, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two highly similar feature requests, but they target different model additions: HelpingAI-3B-v2.2 vs HelpingAIForCausalLM. They are not duplicates of the same underlying change.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #30396 is the clearest standalone representative because its title is more specific about the model being added, but it does not subsume #31103 since they request different artifacts.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #30396 is the best representative for the cluster only in a loose similarity sense, but the items should not be merged as duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:30396", + "right": "issue:31103", + "accept": false, + "reason": "High textual similarity, but the requested additions are different models and therefore different concrete changes; they should be treated as separate issues." + } + ] + }, + "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": 28, + "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-36718-23", + "nodes": [ + "issue:36718", + "issue:36817", + "issue:36879", + "issue:37015", + "issue:37334", + "issue:40330", + "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:37015|issue:37334", + "issue:36879|issue:37015", + "issue:37334|issue:41115", + "issue:37334|issue:43837", + "issue:37334|issue:43671", + "issue:37334|issue:42503", + "issue:40330|issue:42503", + "issue:40330|issue:43671", + "issue:40330|issue:42954", + "issue:40330|issue:46653", + "issue:40330|issue:42738", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41211|issue:43671", + "issue:37015|issue:42503", + "issue:42503|issue:43837", + "issue:42738|issue:45027", + "issue:42503|issue:42738", + "issue:42738|issue:42954", + "issue:42738|issue:46653", + "issue:42954|issue:44205", + "issue:37334|issue:42954", + "issue:42954|issue:43671", + "issue:42503|issue:42954", + "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:41115|issue:45027", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:37334|issue:45027", + "issue:43837|issue:45177", + "issue:42954|issue:45177", + "issue:45322|issue:46133", + "issue:42954|issue:46653", + "issue:42503|issue:46653", + "issue:45027|issue:46653" + ], + "prepared_review_unit_hash": "0a790c59235f4c232afdd8adb3a5bd21eebf1b77903b59672f29d5cbc47f088c", + "component_budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 59, + "serialized_chars": 33618, + "estimated_input_tokens": 8405, + "estimated_eval_tokens": 17066 + }, + "budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 59, + "serialized_chars": 33618, + "estimated_input_tokens": 8405, + "estimated_eval_tokens": 17066 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a set of distinct model-support/model-addition requests that only look similar because they share template wording. I found no true duplicate issue/PR pairs; the soft links are thematic rather than same-change duplicates.", + "confidence": 0.84, + "canonical_issue_reason": "issue:37334 is the most central and broadly phrased request in the set, with several soft links and discussion activity, so it is the best representative issue for this cluster's theme.", + "canonical_pr_reason": null, + "best_issue_reason": "If one issue must stand in for the cluster, issue:37334 is the best fit because it is a broad model-support request and sits near the middle of the similarity graph; still, it should not be treated as a duplicate umbrella for the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36718", + "right": "issue:47004", + "accept": false, + "reason": "Related domain (keypoint matching), but RoMa vs LoMa are different models and different requests." + }, + { + "left": "issue:36718", + "right": "issue:41115", + "accept": false, + "reason": "Different model families and tasks: RoMa keypoint matcher vs MiniCPM3 model architecture." + }, + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "Both are model-add requests, but EuroBert and DEIMv2 are unrelated model families." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "EuroBert config support and Qwen3-TTS support are different model requests." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "EuroBert config vs RF-DETR model support are unrelated additions." + }, + { + "left": "issue:36817", + "right": "issue:42954", + "accept": false, + "reason": "EuroBert and XiaoMi MiMo are different model families with no same underlying change." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "RF-DETR and DEIMv2 are distinct object-detection model requests." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "RF-DETR and LFM2.5 Audio are different modalities and different implementations." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "RF-DETR model support is unrelated to a weight-sparse transformer request." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "RF-DETR and Qwen3-ASR target different model families and code paths." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "RF-DETR vision model support is unrelated to Qwen3-TTS support." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "NeoBERT and Qwen3-ASR are different model additions, not the same change." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "NeoBERT support and weight-sparse transformer support are different implementation requests." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "NeoBERT and LFM2.5 Audio are different model families and tasks." + }, + { + "left": "issue:37015", + "right": "issue:37334", + "accept": false, + "reason": "NeoBERT and LLMDet are distinct model requests; only the request template is similar." + }, + { + "left": "issue:36879", + "right": "issue:37015", + "accept": false, + "reason": "RF-DETR and NeoBERT are unrelated model additions." + }, + { + "left": "issue:37334", + "right": "issue:41115", + "accept": false, + "reason": "LLMDet support and MiniCPM3 architecture support are different models and scopes." + }, + { + "left": "issue:37334", + "right": "issue:43837", + "accept": false, + "reason": "LLMDet support and Qwen3-ASR support are separate model requests." + }, + { + "left": "issue:37334", + "right": "issue:43671", + "accept": false, + "reason": "LLMDet and Qwen3-TTS are different model families and code paths." + }, + { + "left": "issue:37334", + "right": "issue:42503", + "accept": false, + "reason": "LLMDet support and ModernVBERT model support are unrelated additions." + }, + { + "left": "issue:40330", + "right": "issue:42503", + "accept": false, + "reason": "open-qwen2vl-base and ModernVBERT are different models." + }, + { + "left": "issue:40330", + "right": "issue:43671", + "accept": false, + "reason": "open-qwen2vl-base and Qwen3-TTS are unrelated model requests." + }, + { + "left": "issue:40330", + "right": "issue:42954", + "accept": false, + "reason": "open-qwen2vl-base and XiaoMi MiMo are different model families." + }, + { + "left": "issue:40330", + "right": "issue:46653", + "accept": false, + "reason": "open-qwen2vl-base is a model request; Mirko Privitera is unrelated." + }, + { + "left": "issue:40330", + "right": "issue:42738", + "accept": false, + "reason": "open-qwen2vl-base and BERT-like models with RoPE are not the same change." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "MiniCPM3 architecture support and Qwen3-ASR support are different model additions." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "MiniCPM3 and OpenAI weight-sparse transformer are unrelated requests." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "MiniCPM3 architecture support and ModernVBERT support are separate model additions." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "DEIMv2 and Qwen3-TTS are different model families and tasks." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "NeoBERT and ModernVBERT are different model requests." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "ModernVBERT and Qwen3-ASR are unrelated model additions." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "ModernVBERT support and BERT-like models with RoPE are related only at a broad language-model level, not the same underlying change." + }, + { + "left": "issue:42503", + "right": "issue:42954", + "accept": false, + "reason": "ModernVBERT and XiaoMi MiMo are different model families." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Qwen3-TTS and Qwen3-ASR are different concrete support requests." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-TTS and Voxtral-4B-TTS-2603 are different TTS models, not the same bug/change." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-TTS and DeepSeek-OCR2 are different models and different tasks." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-TTS support is unrelated to the weight-sparse transformer request." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Qwen3-ASR and LFM2.5 Audio are different audio-model requests." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Qwen3-ASR and EUPE are different model families and purposes." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Qwen3-ASR support and weight-sparse transformer support are different changes." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "LFM2.5 Audio and SAM3-LiteText are different model-support requests." + }, + { + "left": "issue:43909", + "right": "issue:46645", + "accept": false, + "reason": "LFM2.5 Audio support and Granite SWA+sink options are different code paths." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Weight-sparse transformer support and EUPE are unrelated model requests." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "MiniCPM3 architecture support and Voxtral TTS support are different model additions." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Qwen3-ASR and Voxtral TTS are separate audio-model requests." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "NeoBERT and Voxtral TTS are different model families and tasks." + }, + { + "left": "issue:37334", + "right": "issue:45027", + "accept": false, + "reason": "LLMDet and Voxtral TTS are unrelated support requests." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Qwen3-ASR and DeepSeek-OCR2 are different tasks and models." + }, + { + "left": "issue:42954", + "right": "issue:45177", + "accept": false, + "reason": "XiaoMi MiMo and DeepSeek-OCR2 are unrelated model requests." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "EUPE and TIPSv2 are distinct vision-language model requests." + }, + { + "left": "issue:42954", + "right": "issue:46653", + "accept": false, + "reason": "XiaoMi MiMo and Mirko Privitera are unrelated." + }, + { + "left": "issue:42503", + "right": "issue:46653", + "accept": false, + "reason": "ModernVBERT and Mirko Privitera are unrelated." + }, + { + "left": "issue:45027", + "right": "issue:46653", + "accept": false, + "reason": "Voxtral TTS support and Mirko Privitera are unrelated." + } + ] + }, + "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": 28, + "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": "16fd8b7c26c997f1cfc85b438dc5abf6d92534a539ca69a559fa517723ec4f9a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1506, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1506, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are unrelated: one is a docs request for adding `pruna` to optimization framework docs, while the other is a Vision RoPE refactor in the model code path.", + "confidence": 0.98, + "canonical_issue_reason": "Issue 46443 is the better anchor because it is a concrete technical refactor in the Vision RoPE code path; issue 38740 is a separate documentation request and does not represent the same underlying problem space.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46443 is the strongest standalone issue in this set because it describes a specific implementation change with a clear technical scope, whereas issue 38740 is docs-only and unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Reject: the titles and scopes do not match. One is a documentation feature request, the other is a Vision RoPE refactor; there are no shared files, targets, or evidence of the same underlying 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": 9, + "component_count": 28, + "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": "44c4da4da1a1a051af77626529c66e2d963165de3c347158541571623d878332", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1480, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are not duplicates: one is about block decoding, the other about DeepSeek v3.2 model support.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good canonical for the other because they describe different requests; if forced to pick the more concrete standalone item, 41084 is the narrower technical topic.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different subjects with no shared target or code path: block decoding vs DeepSeek v3.2 support." + } + ] + }, + "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": 28, + "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": "05b05c51f6bbdabb64fef6c81d9c4c202d229cadc54c93eb7aebf76a20cf15f5", + "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 about different features in different model components: one requests a `from_pretrained()` flag for TiledFusedLogitsLoss/LigerFusedLinearCrossEntropy, while the other asks for integrating TiledMLP to reduce memory footprint. They are not duplicate reports of the same bug or change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41306", + "right": "issue:41826", + "accept": false, + "reason": "Different underlying requests: one is about a pretrained-loading flag for FLCE/TiledFusedLogitsLoss, the other is about adding TiledMLP for memory reduction. They touch different functionality and would not be merged as 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": 11, + "component_count": 28, + "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": "34ddaaacba978db0395b335ab6fc33076ba0452d358da06a8d1288ef2876faca", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1584, + "estimated_input_tokens": 396, + "estimated_eval_tokens": 1048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated Trainer customization issues: one about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": "Neither issue is clearly canonical for the other; they address different Trainer behaviors and different code paths.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43598 is the more specific customization concern, but it is not a duplicate of #43600 and should not be treated as a representative for that topic.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Different underlying problems: one concerns unused-column removal/customizability in Trainer, the other concerns FLOPS estimation for custom trainers. Same broad area, but not the same bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 12, + "component_count": 28, + "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": "468f7d5c3d1cea4e7dd16b9c4ae7a6b045b6f7b94c8dad045b3d73bae2b3e522", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1611, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1611, + "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 two different model-support requests in the Transformers registration layer: one for Qwen3-Omni AutoModel/AutoModelForConditionalGeneration, the other for Qwen3.5 AutoModelForSequenceClassification. They do not describe the same underlying bug or change.", + "confidence": 0.96, + "canonical_issue_reason": "Issue 43684 is the broader and more general registration request for Qwen3-Omni, so it is the better representative issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43684 is the stronger global representative because it targets core model registration rather than a narrower downstream sequence-classification variant.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different model families and different APIs/tasks: Qwen3-Omni registration for AutoModel/AutoModelForConditionalGeneration vs Qwen3.5 sequence-classification support. Not the same fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 13, + "component_count": 28, + "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": "9aeb06d848ea67e4010ef9747519442f1c4eacb7309b65f372d4281deb638c65", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4264, + "estimated_input_tokens": 1066, + "estimated_eval_tokens": 2388 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Three issues are near-duplicates with the same truncated title \"huggingface \"; the \"Fett\" issue looks unrelated. I would collapse the three huggingface tickets together and keep the earliest one as canonical.", + "confidence": 0.92, + "canonical_issue_reason": "issue #43845 is the earliest of the near-identical \"huggingface \" issues, so it is the best canonical representative.", + "canonical_pr_reason": null, + "best_issue_reason": "#43845 is the cleanest cluster anchor: same title as the other two matching issues, earliest creation time, and no conflicting signals.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles and no evidence they describe the same underlying problem; \"Fett\" does not match the huggingface cluster." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no shared details beyond weak similarity; insufficient evidence of the same issue." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "\"Fett\" appears unrelated to the huggingface tickets and lacks any concrete overlap." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": true, + "reason": "Same truncated title \"huggingface \" and no distinguishing body details; these look like duplicate reports of the same issue." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": true, + "reason": "Same title and no differing context provided, so they plausibly refer to the same underlying issue." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": true, + "reason": "Same title and no differentiating information; these appear to be duplicates." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: #43845/#43846/#43859 share the exact truncated title and #43845 is the earliest, while #43835 has a distinct title and much weaker similarity. The duplicate/cluster judgments are conservative enough, and nothing appears overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 28, + "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": "cd83e2cbd26e2f5d1b3c61e707b8d206c978c2d863b8c858a2e06168afe1d96d", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7352, + "estimated_input_tokens": 1838, + "estimated_eval_tokens": 3932 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 12, + "serialized_chars": 7352, + "estimated_input_tokens": 1838, + "estimated_eval_tokens": 3932 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The cluster is mostly low-information spam or generic placeholder issues with no convincing shared bug/change. The only exact textual duplicate is the pair 44572/44573 (''), but that is still too content-free to confidently treat as the same underlying issue as the rest.", + "confidence": 0.92, + "canonical_issue_reason": "issue:44572 is the earliest of the only exact duplicate pair ('' / '') and is the cleanest representative among otherwise unrelated, low-signal issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44572 is the best available representative because it matches 44573 exactly by title and is the only plausible duplicate anchor in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are placeholder spam, but there is no evidence they describe the same underlying report." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Both are placeholder spam, but the shared label is too generic to prove a duplicate." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "Spam placeholder vs. generic documentation suggestion; unrelated." + }, + { + "left": "issue:44371", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder vs. benchmark tip; unrelated." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder vs. 'Check out \"Google LLC\"'; no shared bug/change." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "Generic documentation suggestion vs. spam placeholder; not the same issue." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "Generic documentation suggestion vs. spam placeholder; not the same issue." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Exact same placeholder title, but there is still no substantive evidence of a shared underlying bug/change." + }, + { + "left": "issue:44572", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder vs. benchmark tip; unrelated." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder vs. unrelated promotional-looking title; no duplicate evidence." + }, + { + "left": "issue:44573", + "right": "issue:47036", + "accept": false, + "reason": "Spam placeholder vs. benchmark tip; unrelated." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "Spam placeholder vs. unrelated promotional-looking title; no duplicate evidence." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 15, + "component_count": 28, + "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": "bb05c247776f48a207eb4a3f108c796c7b3a30d6c9e33ed66a5d83e1bf3ca51a", + "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": "#45145 and #45151 look like a duplicate issue pair: they have the same title, similar-sized bodies, and a soft-similarity signal. #45145 is the better canonical representative because it has an inbound reference, giving it slightly stronger linkage.", + "confidence": 0.87, + "canonical_issue_reason": "#45145 is the better canonical issue because it has the same title as #45151 and has an inbound reference, so it is the more connected record.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the best issue to keep as the representative duplicate target: it is tied to the same titled report and has the stronger reference signal.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": true, + "reason": "Same title plus the provided similarity signal and comparable body sizes make this a plausible duplicate issue pair, so it is reasonable to treat them as the same request." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The pair is plausibly related, but the rationale overstates the evidence. The packet only shows identical titles, similar body lengths, and a soft-similarity signal; it does not provide the issue bodies, so claims of \u201cnear-identical content\u201d or the same underlying request are not directly grounded. The canonical-choice note about #45145 having an inbound reference is supported, but the duplicate-verdict language should be more conservative." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 16, + "component_count": 28, + "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": 17, + "component_count": 28, + "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": "6be3dd2267072abaa1b6435fdb6bb364479f0ba0a4f52e83fddec91fca0684ab", + "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": "Two closed issues with person-name-and-date titles and no bodies or links; they do not appear to describe the same bug or request.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "Different names and birth dates, with no shared context or evidence of the same underlying issue." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 18, + "component_count": 28, + "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": "f2dc1defd9ac31a3aaf426c826f1fbd2bdc3039f9fc4f59ffb77502a978e17af", + "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 issues do not describe the same bug or request; they look like unrelated name/date entries rather than duplicate technical reports.", + "confidence": 0.98, + "canonical_issue_reason": "No canonical issue: the titles are distinct and contain no shared problem description, repro, or target area.", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good canonical representative because neither contains substantive issue content to anchor deduplication.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "Reject: the titles are different personal-name/date strings with no evidence of the same underlying bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 19, + "component_count": 28, + "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": "54d0c733f1cfdab185583f480a52470e4db181e20a635c16121b782240cdd940", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1541, + "estimated_input_tokens": 386, + "estimated_eval_tokens": 1028 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are unrelated feature requests: one asks for Qwen3_5 token classification/value-model support, the other asks for FP8 training support. They do not describe the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different feature areas and different requested changes; no shared code-path or concrete duplicate signal." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 28, + "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": "6316f67bca0c8a7564076b12585665b9a35f4f67468604d4d29267440b50d985", + "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": "This cluster appears to contain two separate likely-duplicate issue pairs: 45870/45871 and 45872/45873. The evidence shown is strong but not definitive, based mainly on identical titles and matching body lengths. 45869 looks related by domain, but it is a separate dataset promo/request rather than the same underlying issue.", + "confidence": 0.82, + "canonical_issue_reason": "No single canonical issue cleanly covers the whole cluster, since it likely splits into two separate near-duplicate subgroups. If a representative must be chosen for the first subgroup, 45870 is the best anchor because it is the earlier of the near-identical 45870/45871 pair; 45871 looks like a likely duplicate.", + "canonical_pr_reason": null, + "best_issue_reason": "45870 is the best overall issue representative here: it is early, self-contained, and part of a strong duplicate-candidate pair. 45869 is a different promotional dataset request, and 45872/45873 form a separate likely-duplicate subgroup.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Related financial/DataForge context, but a dataset promo/request is not the same underlying issue as an example prediction task." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Same mismatch as 45869/45870: shared domain, but different request type and scope." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Identical title and matching body length make this a strong duplicate candidate." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both mention DataForge/financial time series, but one is an example request and the other is a scraping-service request; different underlying ask." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Same as 45870/45872: related theme, but not the same issue." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Different concrete request even though both are in the same general DataForge area." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Related context only; the issue themes do not match closely enough to be duplicates." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Identical title and matching body length make this a strong duplicate candidate." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The clustering is plausible, but the writeup overstates the evidence. The packet only shows identical titles and matching body lengths for 45870/45871 and 45872/45873, which supports a likely-duplicate reading, not a definitive 'exact duplicate' claim. The summary and canonical reasons should be softened to conservative language (e.g. 'likely duplicate' or 'strong duplicate candidate'). 45869 being separate from the DataForge example pair is grounded, but the duplicate certainty is too strong for the available evidence." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 21, + "component_count": 28, + "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": "074c14a904ec0d17a19b58d32b376d540c6f975ac4d342f8e83f95d5dccee838", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1641, + "estimated_input_tokens": 411, + "estimated_eval_tokens": 1078 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1641, + "estimated_input_tokens": 411, + "estimated_eval_tokens": 1078 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are related to static-cache failures in linear-attention/Qwen3.5 paths, but they describe different failure modes and are not clearly the same underlying bug, so I would not deduplicate them.", + "confidence": 0.74, + "canonical_issue_reason": "Issue #46439 is the better canonical candidate because it states a broader, more root-cause-oriented problem: `early_initialization` corrupts linear-attention layers and leads to shape/runtime failures.", + "canonical_pr_reason": null, + "best_issue_reason": "#46439 best captures the general static-cache/linear-attention corruption issue; #46441 looks like a narrower downstream symptom (`KeyError: 'linear_attention'`) on one model.", + "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 is a shape/runtime corruption bug and the other is a `KeyError` in Qwen3.5. That is too weak to treat as the same concrete underlying bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 22, + "component_count": 28, + "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": "83abc53c47505a1a7dd502b3f09f99e3713afad9dcaf7583ea842f32e7bd06b0", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1709, + "estimated_input_tokens": 428, + "estimated_eval_tokens": 1112 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1709, + "estimated_input_tokens": 428, + "estimated_eval_tokens": 1112 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are exact duplicates: identical titles and apparently identical bodies, describing the same Wav2Vec2CTCTokenizer v5 special-token regression.", + "confidence": 0.99, + "canonical_issue_reason": "issue #46552 is the better canonical issue because it already has inbound references, suggesting it is the more visible/linked duplicate target.", + "canonical_pr_reason": null, + "best_issue_reason": "#46552 is the clearest single issue to keep because it fully matches the bug description and appears to be the more referenced record.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46551", + "right": "issue:46552", + "accept": true, + "reason": "Exact duplicate match: same title and same underlying tokenizer bug report, with no distinguishing details between them." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The decision is well grounded: both issues have the same title, the same body length, and a perfect similarity signal, while #46552 also has inbound references. The duplicate/canonical choice is conservative and consistent with the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 23, + "component_count": 28, + "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": "3f744fdbfdec1f4f8ff4180080ec6d035ea41d61fd41dcc757de037ed8b2f08e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1569, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1569, + "estimated_input_tokens": 393, + "estimated_eval_tokens": 1042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two issues are both about training object detection models, but they describe different failure modes: a tensor size mismatch versus out-of-memory after several epochs. They do not look like the same underlying bug.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #46657 is the more concrete, narrowly defined bug report with a specific tensor-shape mismatch during training, making it a better representative of the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #46657 is the best standalone issue because it describes a specific reproducible error path, unlike #46692 which is a broader resource-exhaustion symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46657", + "right": "issue:46692", + "accept": false, + "reason": "Different underlying problems: one is a tensor dimension mismatch, the other is training-time OOM after some epochs. Similar domain, but not the same bug or change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 24, + "component_count": 28, + "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": "0cb45966323771665dca03a7ad04a6de119c01f0a1e4338935f76998bd219855", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1673, + "estimated_input_tokens": 419, + "estimated_eval_tokens": 1094 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1673, + "estimated_input_tokens": 419, + "estimated_eval_tokens": 1094 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues describe different loading bugs in different code paths: one is ZeRO-3 zero.Init failing to partition a composite submodule and causing OOM, the other is 4-bit + device_map/max_memory incorrectly dispatching a model to CPU and raising ValueError. They should not be deduplicated.", + "confidence": 0.97, + "canonical_issue_reason": "Issue 46822 is the better representative issue for the cluster because it has a clearer, single root-cause description tied to a concrete partitioning failure during multi-GPU load.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46822 is the stronger standalone report: it names the exact affected mechanism (ZeRO-3 zero.Init), the failing submodule, and the observed symptom (OOM), making it more actionable.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46822", + "right": "issue:46823", + "accept": false, + "reason": "Different underlying bugs and code paths: ZeRO-3 zero.Init partitioning/OOM versus 4-bit device_map/max_memory CPU dispatch/ValueError. No shared target or plausible single fix." + } + ] + }, + "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": 28, + "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-46906-23", + "nodes": [ + "issue:46906", + "issue:46940", + "issue:46948", + "issue:46969", + "issue:46985", + "issue:46998", + "issue:47037", + "issue:47049", + "issue:47053", + "issue:47091", + "issue:47102", + "issue:47120", + "issue:47182", + "issue:47227", + "issue:47256", + "issue:47265", + "issue:47279", + "issue:47309", + "issue:47377", + "issue:47390", + "issue:47406", + "issue:47419", + "issue:47442" + ], + "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:46948|issue:46969", + "issue:46948|issue:46985", + "issue:46940|issue:46969", + "issue:46940|issue:46985", + "issue:47037|issue:47049", + "issue:46948|issue:46998", + "issue:47049|issue:47053", + "issue:46998|issue:47049", + "issue:46969|issue:46985", + "issue:46940|issue:46998", + "issue:46985|issue:47049", + "issue:47049|issue:47091", + "issue:46998|issue:47037", + "issue:46985|issue:47037", + "issue:46969|issue:47037", + "issue:47037|issue:47091", + "issue:47037|issue:47053", + "issue:46969|issue:47091", + "issue:47053|issue:47091", + "issue:47091|issue:47102", + "issue:46998|issue:47053", + "issue:47053|issue:47102", + "issue:46985|issue:46998", + "issue:46998|issue:47091", + "issue:47037|issue:47102", + "issue:46985|issue:47091", + "issue:46985|issue:47102", + "issue:47049|issue:47102", + "issue:47265|issue:47279", + "issue:47265|issue:47309", + "issue:47279|issue:47309", + "issue:47265|issue:47377", + "issue:47279|issue:47377", + "issue:47256|issue:47265", + "issue:47256|issue:47279", + "issue:47182|issue:47256", + "issue:47120|issue:47256", + "issue:47309|issue:47377", + "issue:47102|issue:47256", + "issue:47182|issue:47265", + "issue:47120|issue:47182", + "issue:47120|issue:47279", + "issue:47120|issue:47265", + "issue:47120|issue:47377", + "issue:47182|issue:47377", + "issue:47182|issue:47279", + "issue:47227|issue:47377", + "issue:47227|issue:47256", + "issue:47182|issue:47309", + "issue:47227|issue:47279", + "issue:47309|issue:47419", + "issue:47182|issue:47227", + "issue:47227|issue:47265", + "issue:47419|issue:47442", + "issue:47377|issue:47419", + "issue:47377|issue:47442", + "issue:47279|issue:47419", + "issue:47406|issue:47419", + "issue:47279|issue:47442", + "issue:47227|issue:47442", + "issue:47390|issue:47419", + "issue:47390|issue:47406", + "issue:47390|issue:47442", + "issue:47406|issue:47442", + "issue:47377|issue:47406", + "issue:47279|issue:47406" + ], + "prepared_review_unit_hash": "ef2f6af243a632cbfb02d77cd339768de571326a2597740ddd6c932ee7fbbdf8", + "component_budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 74, + "serialized_chars": 39459, + "estimated_input_tokens": 9865, + "estimated_eval_tokens": 19986 + }, + "budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 72, + "serialized_chars": 34125, + "estimated_input_tokens": 8532, + "estimated_eval_tokens": 17320 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a recurring series of dated \"[serge] integration failure triage\" issues, not clear duplicates of one bug. The best anchor is the latest open thread, #47442, which appears to be the active continuation of the series. No PRs are present.", + "confidence": 0.99, + "canonical_issue_reason": "#47442 is the latest open triage thread and has the most current inbound activity, so it is the best anchor for this ongoing series.", + "canonical_pr_reason": null, + "best_issue_reason": "#47442 is the strongest representative of the cluster because it is open, current, and appears to continue the same triage workflow.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46906", + "right": "issue:46940", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46906", + "right": "issue:46948", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46906", + "right": "issue:46969", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46906", + "right": "issue:46985", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46906", + "right": "issue:47037", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46940", + "right": "issue:46948", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46948", + "right": "issue:46969", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46948", + "right": "issue:46985", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46940", + "right": "issue:46969", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46940", + "right": "issue:46985", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47037", + "right": "issue:47049", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46948", + "right": "issue:46998", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47049", + "right": "issue:47053", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46998", + "right": "issue:47049", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46969", + "right": "issue:46985", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46940", + "right": "issue:46998", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46985", + "right": "issue:47049", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47049", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46998", + "right": "issue:47037", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46985", + "right": "issue:47037", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46969", + "right": "issue:47037", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47037", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47037", + "right": "issue:47053", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46969", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47053", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47091", + "right": "issue:47102", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46998", + "right": "issue:47053", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47053", + "right": "issue:47102", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46985", + "right": "issue:46998", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46998", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47037", + "right": "issue:47102", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46985", + "right": "issue:47091", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:46985", + "right": "issue:47102", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47049", + "right": "issue:47102", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47265", + "right": "issue:47279", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47265", + "right": "issue:47309", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47279", + "right": "issue:47309", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47265", + "right": "issue:47377", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47279", + "right": "issue:47377", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47256", + "right": "issue:47265", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47256", + "right": "issue:47279", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47256", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47120", + "right": "issue:47256", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47102", + "right": "issue:47256", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47265", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47120", + "right": "issue:47182", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47120", + "right": "issue:47279", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47120", + "right": "issue:47265", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47120", + "right": "issue:47377", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47377", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47279", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47227", + "right": "issue:47377", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47227", + "right": "issue:47256", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47309", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47227", + "right": "issue:47279", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47309", + "right": "issue:47419", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47182", + "right": "issue:47227", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47227", + "right": "issue:47265", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47419", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47377", + "right": "issue:47419", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47377", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47279", + "right": "issue:47419", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47406", + "right": "issue:47419", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47279", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47227", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47390", + "right": "issue:47419", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47390", + "right": "issue:47406", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47390", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47406", + "right": "issue:47442", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47377", + "right": "issue:47406", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + }, + { + "left": "issue:47279", + "right": "issue:47406", + "accept": false, + "reason": "Same triage-series template, but different dates; no evidence they report the same underlying bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 25, + "component_count": 28, + "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-46906-23", + "nodes": [ + "issue:47227", + "issue:47377", + "issue:47390" + ], + "soft_pairs": [ + "issue:47377|issue:47390", + "issue:47227|issue:47390" + ], + "prepared_review_unit_hash": "a0db6f763ee66bf4762f1df75605b5d5fcf01626596a6a3fea654df2baf28bf9", + "component_budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 74, + "serialized_chars": 39459, + "estimated_input_tokens": 9865, + "estimated_eval_tokens": 19986 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2318, + "estimated_input_tokens": 580, + "estimated_eval_tokens": 1416 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are three date-stamped Serge integration-failure triage issues with very similar templates, but they appear to be separate daily reports rather than duplicates of one underlying bug.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "issue:47377 is the best representative if one is needed: it has the highest inbound references and is a central mid-series triage entry, but it should not be treated as a duplicate of the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47377", + "right": "issue:47390", + "accept": false, + "reason": "Same triage template and subsystem, but the titles and dates indicate separate daily integration-failure reports rather than the same underlying issue." + }, + { + "left": "issue:47227", + "right": "issue:47390", + "accept": false, + "reason": "These look like distinct dated triage notes for different days; similarity is driven by the shared reporting format, not a single bug report." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 26, + "component_count": 28, + "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": "ff7662b123482fb8a86aa89464bd8d78980935f73e70df36e5383ad3fa883d58", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1492, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1492, + "estimated_input_tokens": 373, + "estimated_eval_tokens": 1002 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are near-identical duplicate issues requesting RWKV7 model support; the earlier issue is the best canonical representative.", + "confidence": 0.99, + "canonical_issue_reason": "issue:46983 was opened first and has the same title and apparent request as issue:46984, making it the cleanest canonical issue.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46983 is the strongest single issue to keep because it is earlier and equivalent in scope to issue:46984.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46983", + "right": "issue:46984", + "accept": true, + "reason": "Same title and highly similar body content indicate the same underlying feature request; these are duplicates." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The conclusion is grounded: both issues share the same title, are close in time, and have strong soft-similarity evidence, so treating issue:46983 as the canonical duplicate representative is reasonable and conservative." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 27, + "component_count": 28, + "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": "fc858d2d888b0345003aa6253740706a2b6ac05229c066a8c30504b441c5546c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1626, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1626, + "estimated_input_tokens": 407, + "estimated_eval_tokens": 1070 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items read like standalone article/discussion posts about AI-assisted vulnerability discovery, but they describe different framing and do not point to the same concrete bug, feature, or task. They should not be deduplicated.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue 47015 is the broader, more general title and is the better representative if this cluster is being summarized, but it is not a duplicate of 47007.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47007", + "right": "issue:47015", + "accept": false, + "reason": "The titles and bodies suggest related topic area only; they do not identify the same underlying bug/change or a single concrete target, so this is not a duplicate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 28, + "component_count": 28, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-47138-2", + "nodes": [ + "issue:47138", + "issue:47139" + ], + "soft_pairs": [ + "issue:47138|issue:47139" + ], + "prepared_review_unit_hash": "b92daa53f542c8f2543fc8e13264d525c76f1cbe4a891b68e53f223f0b4cc3ad", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are near-certain duplicate issue reports with the same title and effectively identical content; keep the slightly earlier, more-linked issue as canonical.", + "confidence": 0.99, + "canonical_issue_reason": "Issue #47138 is earlier and has more inbound references, making it the better canonical thread for the duplicate report.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #47138 is the strongest representative because it is first-created and has the higher inbound reference count while describing the same bug as #47139.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:47138", + "right": "issue:47139", + "accept": true, + "reason": "Same title and same reported bug; these are duplicates of the same issue, so they should be clustered together." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative: the two issues have identical titles, near-simultaneous creation times, and strong soft-similarity evidence, while the canonical choice is reasonably supported by earlier creation and higher inbound references." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 31, + "review_unit_index": 1, + "review_unit_count": 5, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-97", + "nodes": [ + "pull_request:43251", + "pull_request:43254", + "pull_request:44775", + "pull_request:46266", + "pull_request:46381", + "pull_request:46836", + "pull_request:46912", + "pull_request:46957", + "pull_request:47077", + "pull_request:47078", + "pull_request:47105", + "pull_request:47108", + "pull_request:47122", + "pull_request:47123", + "pull_request:47124", + "pull_request:47129", + "pull_request:47166", + "pull_request:47173", + "pull_request:47202", + "pull_request:47204", + "pull_request:47212", + "pull_request:47216", + "pull_request:47224", + "pull_request:47243", + "pull_request:47268", + "pull_request:47275", + "pull_request:47280", + "pull_request:47281", + "pull_request:47282", + "pull_request:47284", + "pull_request:47291", + "pull_request:47292", + "pull_request:47294", + "pull_request:47301", + "pull_request:47310", + "pull_request:47360", + "pull_request:47374", + "pull_request:47380", + "pull_request:47384", + "pull_request:47441", + "pull_request:47443", + "pull_request:47444", + "pull_request:47445", + "pull_request:47447", + "pull_request:47449", + "pull_request:47450", + "pull_request:47457", + "pull_request:47462" + ], + "soft_pairs": [ + "pull_request:43251|pull_request:43254", + "pull_request:47282|pull_request:47310", + "pull_request:47124|pull_request:47282", + "pull_request:47280|pull_request:47282", + "pull_request:47282|pull_request:47444", + "pull_request:47123|pull_request:47281", + "pull_request:47124|pull_request:47380", + "pull_request:47122|pull_request:47224", + "pull_request:47122|pull_request:47280", + "pull_request:47124|pull_request:47445", + "pull_request:47310|pull_request:47445", + "pull_request:47282|pull_request:47445", + "pull_request:47124|pull_request:47443", + "pull_request:47281|pull_request:47445", + "pull_request:47268|pull_request:47284", + "pull_request:47173|pull_request:47284", + "pull_request:47284|pull_request:47449", + "pull_request:47202|pull_request:47204", + "pull_request:47204|pull_request:47374", + "pull_request:47284|pull_request:47292", + "pull_request:47108|pull_request:47202", + "pull_request:47077|pull_request:47078", + "pull_request:47224|pull_request:47457", + "pull_request:47204|pull_request:47243", + "pull_request:47202|pull_request:47360", + "pull_request:47108|pull_request:47360", + "pull_request:47292|pull_request:47457", + "pull_request:47441|pull_request:47457", + "pull_request:47105|pull_request:47129", + "pull_request:47449|pull_request:47457", + "pull_request:46266|pull_request:46957", + "pull_request:47077|pull_request:47212", + "pull_request:47301|pull_request:47360", + "pull_request:47173|pull_request:47457", + "pull_request:47360|pull_request:47450", + "pull_request:47360|pull_request:47447", + "pull_request:47384|pull_request:47447", + "pull_request:44775|pull_request:47275", + "pull_request:47301|pull_request:47462", + "pull_request:46957|pull_request:47216", + "pull_request:46381|pull_request:47457", + "pull_request:46836|pull_request:47462", + "pull_request:47166|pull_request:47173", + "pull_request:47291|pull_request:47384", + "pull_request:47216|pull_request:47275", + "pull_request:47216|pull_request:47450", + "pull_request:47077|pull_request:47275", + "pull_request:47294|pull_request:47450", + "pull_request:47275|pull_request:47301", + "pull_request:47216|pull_request:47384", + "pull_request:46912|pull_request:47284" + ], + "prepared_review_unit_hash": "e5ec8d301ea0629a53658f18edb19f26125d2c51f3a508e1bff44896e487d043", + "component_budget": { + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 51, + "serialized_chars": 126888, + "estimated_input_tokens": 31722, + "estimated_eval_tokens": 63700 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous and does not form a single duplicate set. The only clear duplicate pair is the two fixed_cross_entropy PRs; one other edge (EdgeTam original_sizes handling) is plausibly the same bug, but most remaining soft links are false positives across unrelated models, docs, CI, exporter, and quantization changes.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Same underlying fix in src/transformers/loss/loss_utils.py: forward supported kwargs to nn.functional.cross_entropy, including weight and label_smoothing; both target issue 43240." + }, + { + "left": "pull_request:47282", + "right": "pull_request:47310", + "accept": false, + "reason": "Different model/test failures and different code paths: Mamba2 OOM workaround vs Gemma3 output mismatch." + }, + { + "left": "pull_request:47124", + "right": "pull_request:47282", + "accept": false, + "reason": "Both are test-fix PRs, but they address different models and different failure mechanics (Falcon-Mamba output mismatch vs Mamba2 OOM)." + }, + { + "left": "pull_request:47280", + "right": "pull_request:47282", + "accept": false, + "reason": "Same model family, but not the same concrete change: one adds temp offload handling for loading, the other changes generation length/cache cleanup for OOM." + }, + { + "left": "pull_request:47282", + "right": "pull_request:47444", + "accept": false, + "reason": "Unrelated failures in different models: Mamba2 OOM vs MPT load_error." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47281", + "accept": false, + "reason": "Different model integrations and failures: MiniMax-M3-VL load_error vs GLM46V test failures." + }, + { + "left": "pull_request:47124", + "right": "pull_request:47380", + "accept": false, + "reason": "Both touch test stability, but they fix different models and different symptoms; not the same underlying change." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47224", + "accept": true, + "reason": "Same EdgeTam bug: both drop original_sizes before calling the TimmWrapper backbone because the backbone does not accept that kwarg." + }, + { + "left": "pull_request:47122", + "right": "pull_request:47280", + "accept": false, + "reason": "EdgeTam import_or_config failure is unrelated to MiniMax-M3-VL loading/OOM handling." + }, + { + "left": "pull_request:47124", + "right": "pull_request:47445", + "accept": false, + "reason": "Falcon-Mamba output mismatch and PEFT/DeepSeek_v3 loading issues are unrelated." + }, + { + "left": "pull_request:47310", + "right": "pull_request:47445", + "accept": false, + "reason": "Gemma3 generation mismatch vs PEFT/DeepSeek_v3 loading behavior; no shared bug or code path." + }, + { + "left": "pull_request:47282", + "right": "pull_request:47445", + "accept": false, + "reason": "Different models and failures; no common concrete fix." + }, + { + "left": "pull_request:47124", + "right": "pull_request:47443", + "accept": false, + "reason": "Different model families and failure modes; not a duplicate." + }, + { + "left": "pull_request:47281", + "right": "pull_request:47445", + "accept": false, + "reason": "GLM46V integration failures and PEFT/DeepSeek_v3 loading issues are unrelated." + }, + { + "left": "pull_request:47268", + "right": "pull_request:47284", + "accept": false, + "reason": "Different test failures in different models (Zaya vs MIMO v2 Flash)." + }, + { + "left": "pull_request:47173", + "right": "pull_request:47284", + "accept": false, + "reason": "XPU expectation updates across multiple models are unrelated to MIMO v2 Flash test fixes." + }, + { + "left": "pull_request:47284", + "right": "pull_request:47449", + "accept": false, + "reason": "MIMO v2 Flash test fix and Qwen3 Omni Moe test fix are separate model-specific changes." + }, + { + "left": "pull_request:47202", + "right": "pull_request:47204", + "accept": false, + "reason": "Different concerns: deleting CI docker-image builder workflows vs adding exporter test dependencies to torch-light." + }, + { + "left": "pull_request:47204", + "right": "pull_request:47374", + "accept": false, + "reason": "Exporter-test environment setup is unrelated to exporter docs restructuring." + }, + { + "left": "pull_request:47284", + "right": "pull_request:47292", + "accept": false, + "reason": "Different models and failure modes; no shared fix." + }, + { + "left": "pull_request:47108", + "right": "pull_request:47202", + "accept": false, + "reason": "Doc-builder comment cleanup and docker-image builder removal are unrelated changes." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47078", + "accept": false, + "reason": "Both edit Japanese docs, but one is a DistilBERT translation addition while the other is broader wording cleanup; overlap in toctree is incidental." + }, + { + "left": "pull_request:47224", + "right": "pull_request:47457", + "accept": false, + "reason": "EdgeTam import_or_config fix and Inkling test adjustments are unrelated." + }, + { + "left": "pull_request:47204", + "right": "pull_request:47243", + "accept": false, + "reason": "Both concern exporters, but one is dependency/test-environment setup and the other is exporter API/behavior changes; not the same code-path fix." + }, + { + "left": "pull_request:47202", + "right": "pull_request:47360", + "accept": false, + "reason": "Docker builder removal and OTEL CI workflow changes are distinct infrastructure changes." + }, + { + "left": "pull_request:47108", + "right": "pull_request:47360", + "accept": false, + "reason": "Dependabot-friendly doc-builder comments are unrelated to the OTEL CI workflow trim." + }, + { + "left": "pull_request:47292", + "right": "pull_request:47457", + "accept": false, + "reason": "Different models and failure types; no duplicate relationship." + }, + { + "left": "pull_request:47441", + "right": "pull_request:47457", + "accept": false, + "reason": "Whisper dtype/max_length test fixes and Inkling test fixes are unrelated." + }, + { + "left": "pull_request:47105", + "right": "pull_request:47129", + "accept": false, + "reason": "VITS test re-enable and RT-DETR v2 converter fixes are unrelated." + }, + { + "left": "pull_request:47291", + "right": "pull_request:47384", + "accept": false, + "reason": "Deprecating continuous_batching_config and explicit Gemma 4 heterogeneous config handling are separate generation-config changes." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47275", + "accept": false, + "reason": "Compressed-tensors fp8 support and EfficientViT-SAM model addition are unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47275", + "accept": false, + "reason": "Japanese DistilBERT docs and EfficientViT-SAM model addition are unrelated." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47450", + "accept": false, + "reason": "Both touch integration/kernel dispatch helpers, but they implement different features for different model paths and are not the same fix." + }, + { + "left": "pull_request:47275", + "right": "pull_request:47301", + "accept": false, + "reason": "EfficientViT-SAM model addition and MTP docs are unrelated." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47216", + "accept": false, + "reason": "Finegrained FP8 generation/quantizer changes and compressed-tensors fp8 are different quantization features." + }, + { + "left": "pull_request:46381", + "right": "pull_request:47457", + "accept": false, + "reason": "Tokenizers version pinning and Inkling test fixes are unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47462", + "accept": false, + "reason": "Both touch conversion_mapping, but one is Dinov2/DINO-related model conversion and the other is MTP/GLM OCR mapping; different code paths." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47173", + "accept": false, + "reason": "GPTQ quantization tests and XPU expectation updates are unrelated." + }, + { + "left": "pull_request:47291", + "right": "pull_request:47384", + "accept": false, + "reason": "Continuous-batching deprecation and Gemma 4 heterogeneous config changes are different configuration changes." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47450", + "accept": false, + "reason": "Compressed-tensors support and linear-attention kernel dispatch simplification are unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47275", + "accept": false, + "reason": "Doc translation work and new model addition are unrelated." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47450", + "accept": false, + "reason": "Kernel-native causal_conv1d support and linear-attention dispatch simplification are different features, not one duplicate change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 31, + "review_unit_index": 2, + "review_unit_count": 5, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-97", + "nodes": [ + "pull_request:43451", + "pull_request:44517", + "pull_request:44775", + "pull_request:45157", + "pull_request:46196", + "pull_request:46266", + "pull_request:46395", + "pull_request:46509", + "pull_request:46604", + "pull_request:46658", + "pull_request:46836", + "pull_request:46912", + "pull_request:46995", + "pull_request:47003", + "pull_request:47077", + "pull_request:47078", + "pull_request:47082", + "pull_request:47093", + "pull_request:47108", + "pull_request:47113", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47175", + "pull_request:47179", + "pull_request:47202", + "pull_request:47204", + "pull_request:47210", + "pull_request:47216", + "pull_request:47243", + "pull_request:47275", + "pull_request:47292", + "pull_request:47294", + "pull_request:47295", + "pull_request:47301", + "pull_request:47304", + "pull_request:47320", + "pull_request:47336", + "pull_request:47357", + "pull_request:47384", + "pull_request:47422", + "pull_request:47447", + "pull_request:47449", + "pull_request:47450", + "pull_request:47451", + "pull_request:47452", + "pull_request:47461", + "pull_request:47462" + ], + "soft_pairs": [ + "pull_request:46912|pull_request:47179", + "pull_request:46196|pull_request:47202", + "pull_request:47204|pull_request:47357", + "pull_request:46196|pull_request:47210", + "pull_request:47166|pull_request:47449", + "pull_request:47003|pull_request:47204", + "pull_request:46836|pull_request:47202", + "pull_request:47093|pull_request:47275", + "pull_request:47082|pull_request:47462", + "pull_request:45157|pull_request:47093", + "pull_request:46995|pull_request:47163", + "pull_request:45157|pull_request:46395", + "pull_request:46836|pull_request:47275", + "pull_request:46266|pull_request:47462", + "pull_request:47204|pull_request:47295", + "pull_request:47216|pull_request:47462", + "pull_request:46912|pull_request:47082", + "pull_request:47163|pull_request:47384", + "pull_request:44775|pull_request:47301", + "pull_request:46196|pull_request:46995", + "pull_request:47301|pull_request:47447", + "pull_request:47447|pull_request:47462", + "pull_request:47447|pull_request:47450", + "pull_request:47301|pull_request:47384", + "pull_request:47243|pull_request:47304", + "pull_request:47077|pull_request:47301", + "pull_request:47243|pull_request:47447", + "pull_request:46604|pull_request:47357", + "pull_request:47163|pull_request:47451", + "pull_request:44775|pull_request:47077", + "pull_request:47082|pull_request:47216", + "pull_request:47166|pull_request:47292", + "pull_request:47294|pull_request:47320", + "pull_request:47113|pull_request:47422", + "pull_request:47163|pull_request:47357", + "pull_request:46196|pull_request:47179", + "pull_request:47093|pull_request:47143", + "pull_request:46509|pull_request:46995", + "pull_request:46995|pull_request:47143", + "pull_request:46266|pull_request:47301", + "pull_request:47450|pull_request:47462", + "pull_request:44775|pull_request:47304", + "pull_request:47093|pull_request:47163", + "pull_request:47082|pull_request:47210", + "pull_request:47179|pull_request:47210", + "pull_request:46509|pull_request:47357", + "pull_request:46604|pull_request:47003", + "pull_request:47082|pull_request:47108", + "pull_request:47078|pull_request:47275", + "pull_request:47108|pull_request:47301", + "pull_request:43451|pull_request:47357", + "pull_request:46266|pull_request:47216", + "pull_request:46266|pull_request:47336", + "pull_request:46995|pull_request:47301", + "pull_request:47175|pull_request:47179", + "pull_request:44775|pull_request:46836", + "pull_request:46604|pull_request:46658", + "pull_request:47163|pull_request:47179", + "pull_request:47320|pull_request:47452", + "pull_request:44517|pull_request:45157", + "pull_request:47275|pull_request:47336", + "pull_request:47210|pull_request:47304", + "pull_request:46196|pull_request:47461", + "pull_request:47204|pull_request:47422", + "pull_request:47113|pull_request:47452", + "pull_request:47336|pull_request:47357", + "pull_request:46912|pull_request:47078", + "pull_request:46196|pull_request:47450", + "pull_request:47108|pull_request:47204" + ], + "prepared_review_unit_hash": "084c3cca603af894aca73cfe5ca1f3b6e2e05d18f09a83af15f88fd2d1ef87e3", + "component_budget": { + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 69, + "serialized_chars": 141333, + "estimated_input_tokens": 35334, + "estimated_eval_tokens": 70924 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous: it mixes unrelated model additions, docs-only updates, CI/docker changes, exporter work, attention/backend fixes, and test tweaks. None of the soft pairs looks like the same underlying bug or change, so I would not collapse any of them as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single PR is a safe canonical representative; the items span unrelated features and fixes across models, docs, CI, and export/runtime plumbing.", + "best_issue_reason": null, + "best_pr_reason": "No global best PR for deduping: the cluster is not one coherent change, so picking any PR would be arbitrary.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46912", + "right": "pull_request:47179", + "accept": false, + "reason": "Both are RoPE-related only in the broadest sense, but one is test coverage and the other is new Granite SWA model support; different changes." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47202", + "accept": false, + "reason": "One is SDPA dispatch for audio models, the other is removing a CI docker builder; unrelated code paths." + }, + { + "left": "pull_request:47204", + "right": "pull_request:47357", + "accept": false, + "reason": "Exporter test-image dependency changes versus FSDP CI/save logic; not the same bug or feature." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47210", + "accept": false, + "reason": "Both touch attention internals, but one is model dispatch support and the other is packed-row varlen attention; different fixes." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47449", + "accept": false, + "reason": "Quantization test updates versus qwen3_omni_moe model test fixes; unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47204", + "accept": false, + "reason": "Both are exporter/tooling related, but one adds OpenVINO exporter support and the other changes CI image dependencies." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47202", + "accept": false, + "reason": "Unlimited OCR model support versus docker image builder removal; unrelated." + }, + { + "left": "pull_request:47093", + "right": "pull_request:47275", + "accept": false, + "reason": "Different new models, different architectures, no shared concrete change." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47462", + "accept": false, + "reason": "RoMa keypoint matcher support versus MTP/GLM OCR generation changes; unrelated." + }, + { + "left": "pull_request:45157", + "right": "pull_request:47093", + "accept": false, + "reason": "PrismML GGUF support versus OpenPangu-V2 model addition; different subsystems." + }, + { + "left": "pull_request:46995", + "right": "pull_request:47163", + "accept": false, + "reason": "A backend availability helper change versus XPU test additions; not the same bug." + }, + { + "left": "pull_request:45157", + "right": "pull_request:46395", + "accept": false, + "reason": "Prism GGUF support versus Param2MoE model support; unrelated." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47275", + "accept": false, + "reason": "Two different model additions with no shared code path." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47462", + "accept": false, + "reason": "DINO conversion/modernization work versus MTP generation support; unrelated changes." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47082", + "accept": false, + "reason": "RoPE test coverage versus a keypoint matcher model addition; not duplicate work." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47384", + "accept": false, + "reason": "Paged-attention XPU tests versus Gemma4 heterogeneity config changes; different issues." + }, + { + "left": "pull_request:44775", + "right": "pull_request:47301", + "accept": false, + "reason": "Docs navigation/parallelism updates versus assisted-decoding MTP docs; different documentation topics." + }, + { + "left": "pull_request:46196", + "right": "pull_request:46995", + "accept": false, + "reason": "Audio SDPA dispatch versus flash-linear-attention availability on MLU; different backend problems." + }, + { + "left": "pull_request:47301", + "right": "pull_request:47447", + "accept": false, + "reason": "Assisted decoding docs versus cache rollback behavior for speculative decoding; not the same change." + }, + { + "left": "pull_request:47301", + "right": "pull_request:47384", + "accept": false, + "reason": "MTP documentation versus Gemma4 heterogeneous config implementation; unrelated." + }, + { + "left": "pull_request:47447", + "right": "pull_request:47462", + "accept": false, + "reason": "Cache rollback for speculative decoding versus MTP for GLM OCR; different generation mechanisms." + }, + { + "left": "pull_request:47447", + "right": "pull_request:47450", + "accept": false, + "reason": "Speculative-decoding cache changes versus linear-attention dispatch simplification; unrelated." + }, + { + "left": "pull_request:47301", + "right": "pull_request:47422", + "accept": false, + "reason": "MTP docs versus CI/docker pinning and RNNT fixes; not duplicate." + }, + { + "left": "pull_request:47243", + "right": "pull_request:47304", + "accept": false, + "reason": "Executorch exporter fixes versus MPS graph-cache docs cleanup; unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47301", + "accept": false, + "reason": "Japanese docs translation updates versus English MTP docs; different documentation work." + }, + { + "left": "pull_request:47243", + "right": "pull_request:47447", + "accept": false, + "reason": "Executorch exporter fixes versus cache rollback support; different code paths." + }, + { + "left": "pull_request:46604", + "right": "pull_request:47357", + "accept": false, + "reason": "Mistral tekken tokenizer utilities versus FSDP CI/save changes; unrelated." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47451", + "accept": false, + "reason": "XPU paged-attention tests versus attention-interface value padding; different problems." + }, + { + "left": "pull_request:44775", + "right": "pull_request:47077", + "accept": false, + "reason": "English distributed-training docs versus Japanese DistilBERT docs translation; unrelated." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47216", + "accept": false, + "reason": "RoMa model support versus compressed-tensors FP8 quantization support; different features." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47292", + "accept": false, + "reason": "Quantization test edits versus TipsV2 DPT XPU test fixes; unrelated." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47320", + "accept": false, + "reason": "Kernel-native causal-conv work versus response-schema deprecation; different areas." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47422", + "accept": false, + "reason": "Audio feature/proc requirement fixes versus CI image torch pinning and RNNT break fix; unrelated." + }, + { + "left": "pull_request:47320", + "right": "pull_request:47452", + "accept": false, + "reason": "Response-schema deprecation versus LFM2/Mamba-style model alignment work; not the same change." + }, + { + "left": "pull_request:44517", + "right": "pull_request:45157", + "accept": false, + "reason": "Qwen3-TTS model support versus Prism GGUF dequantization support; unrelated." + }, + { + "left": "pull_request:47275", + "right": "pull_request:47336", + "accept": false, + "reason": "EfficientViT-SAM model support versus offloaded model saving bugfix; unrelated." + }, + { + "left": "pull_request:47210", + "right": "pull_request:47304", + "accept": false, + "reason": "Packed SDPA runtime change versus MPS graph-cache docs cleanup; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47461", + "accept": false, + "reason": "Audio SDPA dispatch versus Wav2Vec2Phoneme tokenizer language-state fix; different code paths." + }, + { + "left": "pull_request:47204", + "right": "pull_request:47422", + "accept": false, + "reason": "CI/exporter dependencies versus torch pinning in quantization docker; related to CI, but not the same concrete issue." + }, + { + "left": "pull_request:47336", + "right": "pull_request:47357", + "accept": false, + "reason": "Fully offloaded save bugfix versus FSDP CI/end-to-end tests; different problems." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47078", + "accept": false, + "reason": "RoPE testing updates versus Japanese documentation wording improvements; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47450", + "accept": false, + "reason": "Audio SDPA dispatch versus linear-attention function dispatch simplification; different attention backends and code paths." + }, + { + "left": "pull_request:47093", + "right": "pull_request:47163", + "accept": false, + "reason": "OpenPangu-V2 model support versus XPU paged-attention tests; unrelated." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47210", + "accept": false, + "reason": "RoMa model addition versus packed SDPA varlen attention; unrelated." + }, + { + "left": "pull_request:47179", + "right": "pull_request:47210", + "accept": false, + "reason": "Granite SWA model support versus packed SDPA runtime change; not duplicate work." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47357", + "accept": false, + "reason": "Nemotron Omni model support versus FSDP CI/save changes; unrelated." + }, + { + "left": "pull_request:46604", + "right": "pull_request:47003", + "accept": false, + "reason": "Mistral tokenizer conversion utilities versus OpenVINO exporter support; both tooling-related but different concrete features." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47108", + "accept": false, + "reason": "Model support versus doc-builder comment tweak; unrelated." + }, + { + "left": "pull_request:47078", + "right": "pull_request:47275", + "accept": false, + "reason": "Japanese docs wording changes versus EfficientViT-SAM model support; unrelated." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47357", + "accept": false, + "reason": "XPU paged-attention tests versus FSDP CI/save work; different issues." + }, + { + "left": "pull_request:47082", + "right": "pull_request:47210", + "accept": false, + "reason": "RoMa keypoint matcher support versus packed SDPA varlen routing; unrelated." + }, + { + "left": "pull_request:46196", + "right": "pull_request:47461", + "accept": false, + "reason": "Audio SDPA dispatch versus stale phonemizer language fix; different subsystems." + }, + { + "left": "pull_request:47204", + "right": "pull_request:47422", + "accept": false, + "reason": "Exporter/CI image dependency changes versus quantization docker torch pinning; related only at the CI level, not the same bug." + }, + { + "left": "pull_request:47108", + "right": "pull_request:47204", + "accept": false, + "reason": "Doc-builder comment update versus docker exporter-test dependency additions; 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": 31, + "review_unit_index": 3, + "review_unit_count": 5, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-97", + "nodes": [ + "pull_request:41855", + "pull_request:43424", + "pull_request:43451", + "pull_request:44517", + "pull_request:45157", + "pull_request:45273", + "pull_request:46196", + "pull_request:46266", + "pull_request:46381", + "pull_request:46556", + "pull_request:46604", + "pull_request:46658", + "pull_request:46836", + "pull_request:46912", + "pull_request:46957", + "pull_request:47077", + "pull_request:47078", + "pull_request:47096", + "pull_request:47105", + "pull_request:47113", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47175", + "pull_request:47179", + "pull_request:47210", + "pull_request:47216", + "pull_request:47243", + "pull_request:47275", + "pull_request:47291", + "pull_request:47294", + "pull_request:47298", + "pull_request:47301", + "pull_request:47304", + "pull_request:47306", + "pull_request:47320", + "pull_request:47322", + "pull_request:47336", + "pull_request:47360", + "pull_request:47367", + "pull_request:47422", + "pull_request:47438", + "pull_request:47447", + "pull_request:47451", + "pull_request:47452", + "pull_request:47457", + "pull_request:47461", + "pull_request:47462" + ], + "soft_pairs": [ + "pull_request:47360|pull_request:47422", + "pull_request:47336|pull_request:47447", + "pull_request:47179|pull_request:47301", + "pull_request:47175|pull_request:47461", + "pull_request:47322|pull_request:47451", + "pull_request:47291|pull_request:47451", + "pull_request:46556|pull_request:46836", + "pull_request:47216|pull_request:47336", + "pull_request:47166|pull_request:47422", + "pull_request:47275|pull_request:47422", + "pull_request:47298|pull_request:47438", + "pull_request:46912|pull_request:47294", + "pull_request:47143|pull_request:47320", + "pull_request:47175|pull_request:47294", + "pull_request:47175|pull_request:47438", + "pull_request:47216|pull_request:47243", + "pull_request:47077|pull_request:47461", + "pull_request:45273|pull_request:47451", + "pull_request:47216|pull_request:47294", + "pull_request:47078|pull_request:47452", + "pull_request:47175|pull_request:47447", + "pull_request:46912|pull_request:47438", + "pull_request:47105|pull_request:47298", + "pull_request:47078|pull_request:47322", + "pull_request:47322|pull_request:47461", + "pull_request:45157|pull_request:46658", + "pull_request:44517|pull_request:47301", + "pull_request:47294|pull_request:47438", + "pull_request:47113|pull_request:47304", + "pull_request:46912|pull_request:47291", + "pull_request:47438|pull_request:47447", + "pull_request:43451|pull_request:44517", + "pull_request:45273|pull_request:46196", + "pull_request:46556|pull_request:47301", + "pull_request:44517|pull_request:47143", + "pull_request:44517|pull_request:47163", + "pull_request:47210|pull_request:47291", + "pull_request:46604|pull_request:47163", + "pull_request:47113|pull_request:47175", + "pull_request:43451|pull_request:47451", + "pull_request:43451|pull_request:46556", + "pull_request:46556|pull_request:47096", + "pull_request:46957|pull_request:47360", + "pull_request:45273|pull_request:46556", + "pull_request:46381|pull_request:47322", + "pull_request:46957|pull_request:47301", + "pull_request:43451|pull_request:46604", + "pull_request:47298|pull_request:47447", + "pull_request:47113|pull_request:47275", + "pull_request:47077|pull_request:47367", + "pull_request:45273|pull_request:47447", + "pull_request:46266|pull_request:47452", + "pull_request:45273|pull_request:47294", + "pull_request:47294|pull_request:47306", + "pull_request:47210|pull_request:47367", + "pull_request:47275|pull_request:47367", + "pull_request:46957|pull_request:47336", + "pull_request:43451|pull_request:46381", + "pull_request:46381|pull_request:47096", + "pull_request:47166|pull_request:47457", + "pull_request:45273|pull_request:47243", + "pull_request:47322|pull_request:47462", + "pull_request:41855|pull_request:43424" + ], + "prepared_review_unit_hash": "a011fd171f8e6d0945ffaa65bd93cd3cc82878cdc238f87b436266aecfb8b608", + "component_budget": { + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 63, + "serialized_chars": 137099, + "estimated_input_tokens": 34275, + "estimated_eval_tokens": 68806 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is mostly a set of unrelated PRs with superficial similarity from shared subsystems like model additions, export, generation, docs, and CI. None of the soft-edge pairs look like the same underlying bug or change, so I would not merge any of them as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:47360", + "right": "pull_request:47422", + "accept": false, + "reason": "Different CI concerns: OTEL workflow instrumentation vs torch/torchaudio pinning and quantization image fixes." + }, + { + "left": "pull_request:47336", + "right": "pull_request:47447", + "accept": false, + "reason": "One fixes fully offloaded model saving; the other adds sliding-window cache rollback for speculative decoding. Different code paths." + }, + { + "left": "pull_request:47179", + "right": "pull_request:47301", + "accept": false, + "reason": "New Granite-SWA model support is unrelated to assisted decoding docs for multi-token prediction." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47461", + "accept": false, + "reason": "CI expectation updates for Olmo3 are unrelated to the stale phonemizer language bug in Wav2Vec2Phoneme." + }, + { + "left": "pull_request:47322", + "right": "pull_request:47451", + "accept": false, + "reason": "Quark config/version handling is unrelated to moving value padding into attention interfaces." + }, + { + "left": "pull_request:47291", + "right": "pull_request:47451", + "accept": false, + "reason": "Deprecating ContinuousBatchingConfig from generation config is a different change from attention value padding." + }, + { + "left": "pull_request:46556", + "right": "pull_request:46836", + "accept": false, + "reason": "Image size utility support for PIL images is not the same change as adding Unlimited OCR model support." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47336", + "accept": false, + "reason": "Compressed-tensors FP8 integration is unrelated to offloaded model saving behavior." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47422", + "accept": false, + "reason": "Quantization test tweaks for GPTQ/MXFP4 are unrelated to the CI torch pinning fix." + }, + { + "left": "pull_request:47275", + "right": "pull_request:47422", + "accept": false, + "reason": "Adding EfficientVit-SAM is a new model PR, not the same as fixing CI torch/torchaudio versioning." + }, + { + "left": "pull_request:47298", + "right": "pull_request:47438", + "accept": false, + "reason": "MI450 model failures and less-materialized masks both touch runtime behavior, but the concrete problems and fixes are different." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47294", + "accept": false, + "reason": "RoPE test coverage and kernel-native causal_conv1d functions are separate infrastructure changes." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47320", + "accept": false, + "reason": "NeuCodec model support is unrelated to deprecating the old response schema in CLI serving." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47294", + "accept": false, + "reason": "Olmo3 CI output updates do not match the causal_conv1d kernel integration work." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47438", + "accept": false, + "reason": "CI expectation updates for a model test are not the same as boolean mask materialization changes." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47243", + "accept": false, + "reason": "Compressed-tensors FP8 work and Executorch exporter fixes are both infra-related but fix different mechanisms." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47461", + "accept": false, + "reason": "Japanese docs translation edits are unrelated to the Wav2Vec2Phoneme language-state bug." + }, + { + "left": "pull_request:45273", + "right": "pull_request:47451", + "accept": false, + "reason": "Liger eval OOM fix in Trainer is unrelated to attention interface value padding." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47294", + "accept": false, + "reason": "Compressed-tensors integration is unrelated to kernel-native causal_conv1d support." + }, + { + "left": "pull_request:47078", + "right": "pull_request:47452", + "accept": false, + "reason": "Japanese wording/doc updates are unrelated to the broader Mamba alignment refactor." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47447", + "accept": false, + "reason": "Model CI expectation churn and speculative-decoding cache rollback are different issues." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47438", + "accept": false, + "reason": "RoPE test coverage and less-materialized masks are different attention-related changes, not one bug." + }, + { + "left": "pull_request:47105", + "right": "pull_request:47298", + "accept": false, + "reason": "Re-enabling a VITS test is unrelated to MI450 hardware model failures." + }, + { + "left": "pull_request:47078", + "right": "pull_request:47322", + "accept": false, + "reason": "Japanese doc wording is unrelated to Quark config class changes." + }, + { + "left": "pull_request:47322", + "right": "pull_request:47461", + "accept": false, + "reason": "Quark quantization config changes are unrelated to the stale phonemizer language fix." + }, + { + "left": "pull_request:45157", + "right": "pull_request:46658", + "accept": false, + "reason": "Both are large model-support efforts, but they target different architectures and codepaths with no shared bug/change." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47301", + "accept": false, + "reason": "Qwen3-TTS model support is unrelated to assisted-decoding docs for MTP." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47438", + "accept": false, + "reason": "Kernel-native causal_conv1d support and less-materialized masks are distinct optimization paths." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47243", + "accept": false, + "reason": "Audio feature/proc requirement decorators are unrelated to Executorch exporter fixes." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47143", + "accept": false, + "reason": "Qwen3-TTS support and NeuCodec support are different model additions." + }, + { + "left": "pull_request:44517", + "right": "pull_request:47163", + "accept": false, + "reason": "Qwen3-TTS model support is unrelated to XPU paged-attention test coverage." + }, + { + "left": "pull_request:47210", + "right": "pull_request:47291", + "accept": false, + "reason": "Packed SDPA varlen routing and generation config deprecation are unrelated." + }, + { + "left": "pull_request:46604", + "right": "pull_request:47163", + "accept": false, + "reason": "Mistral tekken tokenizer utilities are unrelated to XPU paged-attention tests." + }, + { + "left": "pull_request:47113", + "right": "pull_request:47175", + "accept": false, + "reason": "Audio requirement fixes do not match Olmo3 CI expectation updates." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47451", + "accept": false, + "reason": "Molmo2 model support and attention value padding are completely different changes." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46556", + "accept": false, + "reason": "Molmo2 model support is unrelated to PIL image size handling in image_utils." + }, + { + "left": "pull_request:46556", + "right": "pull_request:47096", + "accept": false, + "reason": "Image utility changes are unrelated to Exportable Kimi exporter support." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47360", + "accept": false, + "reason": "Finegrained FP8 fused-kernel logic is unrelated to the OTEL CI workflow change." + }, + { + "left": "pull_request:45273", + "right": "pull_request:47447", + "accept": false, + "reason": "Trainer eval OOM fix and cache rollback for speculative decoding are different generation/training paths." + }, + { + "left": "pull_request:46266", + "right": "pull_request:47452", + "accept": false, + "reason": "DINO modernization work and Mamba coverage refactors are separate model-family changes." + }, + { + "left": "pull_request:45273", + "right": "pull_request:47294", + "accept": false, + "reason": "Trainer eval logits handling and causal_conv1d kernel integration are unrelated." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47306", + "accept": false, + "reason": "Kernel-native causal_conv1d functions and Neuron experiment scripts are unrelated." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47367", + "accept": false, + "reason": "Japanese translation edits are unrelated to pipeline device-placement clarification." + }, + { + "left": "pull_request:46957", + "right": "pull_request:47336", + "accept": false, + "reason": "Finegrained FP8 decoding optimizations and fully offloaded model saving are unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46381", + "accept": false, + "reason": "Molmo2 model support is unrelated to allowing tokenizers 0.23.1." + }, + { + "left": "pull_request:46381", + "right": "pull_request:47096", + "accept": false, + "reason": "Tokenizer dependency/version compatibility changes are unrelated to Exportable Kimi." + }, + { + "left": "pull_request:47166", + "right": "pull_request:47457", + "accept": false, + "reason": "Quantization test extension and Inkling test fixes are different test suites for different subsystems." + }, + { + "left": "pull_request:45273", + "right": "pull_request:47243", + "accept": false, + "reason": "Liger eval OOM behavior and Executorch exporter fixes are not the same code-path problem." + }, + { + "left": "pull_request:47322", + "right": "pull_request:47462", + "accept": false, + "reason": "Quark config changes and MTP support for GLM OCR/MROPEs are unrelated features." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "Mistral missing tokenizer methods and Executorch dynamic-shape tests do not fix the same problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 31, + "review_unit_index": 4, + "review_unit_count": 5, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-97", + "nodes": [ + "pull_request:41855", + "pull_request:42668", + "pull_request:43251", + "pull_request:43424", + "pull_request:43823", + "pull_request:43995", + "pull_request:44827", + "pull_request:45157", + "pull_request:45170", + "pull_request:45213", + "pull_request:45655", + "pull_request:46048", + "pull_request:46509", + "pull_request:46658", + "pull_request:46687", + "pull_request:46836", + "pull_request:46912", + "pull_request:47003", + "pull_request:47077", + "pull_request:47122", + "pull_request:47123", + "pull_request:47124", + "pull_request:47143", + "pull_request:47163", + "pull_request:47166", + "pull_request:47173", + "pull_request:47175", + "pull_request:47210", + "pull_request:47216", + "pull_request:47243", + "pull_request:47266", + "pull_request:47280", + "pull_request:47281", + "pull_request:47282", + "pull_request:47291", + "pull_request:47292", + "pull_request:47301", + "pull_request:47304", + "pull_request:47310", + "pull_request:47322", + "pull_request:47336", + "pull_request:47380", + "pull_request:47387", + "pull_request:47443", + "pull_request:47444", + "pull_request:47445", + "pull_request:47449", + "pull_request:47461" + ], + "soft_pairs": [ + "pull_request:41855|pull_request:43823", + "pull_request:42668|pull_request:43823", + "pull_request:43424|pull_request:47301", + "pull_request:43823|pull_request:45655", + "pull_request:44827|pull_request:45170", + "pull_request:43424|pull_request:44827", + "pull_request:43823|pull_request:43995", + "pull_request:43251|pull_request:43823", + "pull_request:44827|pull_request:46048", + "pull_request:43424|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:44827|pull_request:45655", + "pull_request:43995|pull_request:45213", + "pull_request:44827|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:43823|pull_request:45213", + "pull_request:46687|pull_request:47461", + "pull_request:47123|pull_request:47280", + "pull_request:47266|pull_request:47443", + "pull_request:47282|pull_request:47380", + "pull_request:47123|pull_request:47443", + "pull_request:47123|pull_request:47266", + "pull_request:47266|pull_request:47280", + "pull_request:47280|pull_request:47443", + "pull_request:47124|pull_request:47310", + "pull_request:47280|pull_request:47380", + "pull_request:47443|pull_request:47445", + "pull_request:47280|pull_request:47281", + "pull_request:47443|pull_request:47444", + "pull_request:47310|pull_request:47380", + "pull_request:47280|pull_request:47310", + "pull_request:47380|pull_request:47444", + "pull_request:47122|pull_request:47281", + "pull_request:47266|pull_request:47444", + "pull_request:47122|pull_request:47123", + "pull_request:47444|pull_request:47445", + "pull_request:47266|pull_request:47445", + "pull_request:47280|pull_request:47444", + "pull_request:47123|pull_request:47444", + "pull_request:47281|pull_request:47443", + "pull_request:47122|pull_request:47310", + "pull_request:47173|pull_request:47449", + "pull_request:47292|pull_request:47449", + "pull_request:47173|pull_request:47292", + "pull_request:46912|pull_request:47175", + "pull_request:47301|pull_request:47304", + "pull_request:47216|pull_request:47301", + "pull_request:47077|pull_request:47304", + "pull_request:46658|pull_request:46836", + "pull_request:47210|pull_request:47301", + "pull_request:47163|pull_request:47166", + "pull_request:47003|pull_request:47243", + "pull_request:46509|pull_request:47143", + "pull_request:46509|pull_request:47163", + "pull_request:45157|pull_request:46836", + "pull_request:47291|pull_request:47322", + "pull_request:46836|pull_request:47387", + "pull_request:45157|pull_request:46509", + "pull_request:47301|pull_request:47336" + ], + "prepared_review_unit_hash": "c6f570b00e017e0f1ae191eebff1baf90d3d8c998d5141191d682b9b29e5ef88", + "component_budget": { + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 59, + "serialized_chars": 128019, + "estimated_input_tokens": 32005, + "estimated_eval_tokens": 64266 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose cluster of mostly unrelated PRs. The only clear duplicate groups are the Wav2Vec2Phoneme tokenizer stale-language fix, the MiniMax-M3-VL OOM/offload test fix, the Mamba2 OOM test fix, and the DeepSeek v3/v32 code-cleanup trio. Most other pairs are different features, models, or docs work and should stay separate.", + "confidence": 0.64, + "canonical_issue_reason": null, + "canonical_pr_reason": "47443 is the best canonical PR anchor because it contains the core shared DeepSeek v3/v32 model-code change that the other true duplicate PRs mirror, while the remaining accepted groups are smaller follow-up test fixes.", + "best_issue_reason": null, + "best_pr_reason": "47443 is the strongest overall representative of the genuine duplicate subcluster: it carries the shared code-path fix across multiple model files and is more central than the smaller test-only variants.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Unrelated topics: Mistral tokenizer API gap vs a new MobileLLM model addition." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor robustness fix and MobileLLM model addition are different changes." + }, + { + "left": "pull_request:43424", + "right": "pull_request:47301", + "accept": false, + "reason": "ExecuTorch dynamic-shape test coverage and MTP docs are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "New model support vs config-resolution ordering fix; different code paths." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test fixes and CLIP-like layernorm rename are unrelated." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "ExecuTorch export test and Mistral4 test fixes are different model/runtime issues." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "Model addition and Falcon output-interface refactor are not the same bug/change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "cross_entropy kwargs passthrough and MobileLLM support are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 test work and TDT loss kernel support do not share a concrete code path." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "ExecuTorch export tests and MobileLLM model addition are unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "Loss-function kwargs fix is unrelated to Mistral4 test failures." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "These touch different subsystems and different failure modes." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactor and model-creation skill docs are not the same change." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "Model tests and a repo skill scaffold are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model support and skill/docs scaffolding are different work." + }, + { + "left": "pull_request:45157", + "right": "pull_request:46836", + "accept": false, + "reason": "PrismML Bonsai support and Unlimited OCR are distinct model additions." + }, + { + "left": "pull_request:46687", + "right": "pull_request:47461", + "accept": true, + "reason": "Same tokenizer bug in the same file; both fix stale phonemizer_lang state and can be merged together." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47280", + "accept": true, + "reason": "Same MiniMax-M3-VL load_error/offload-folder test fix in the same test file." + }, + { + "left": "pull_request:47266", + "right": "pull_request:47443", + "accept": false, + "reason": "Different model families and different fix styles; not the same concrete bug/change." + }, + { + "left": "pull_request:47282", + "right": "pull_request:47380", + "accept": true, + "reason": "Same Mamba2 OOM test issue in the same integration test, with only minor wording/setup differences." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47443", + "accept": false, + "reason": "MiniMax-M3-VL load_error and DeepSeek/MPT refactoring are unrelated." + }, + { + "left": "pull_request:47123", + "right": "pull_request:47266", + "accept": false, + "reason": "Different models and different failing tests; not mergeable as one fix." + }, + { + "left": "pull_request:47266", + "right": "pull_request:47280", + "accept": false, + "reason": "DeepSeek V3.2 load_error and MiniMax-M3-VL load_error are separate model issues." + }, + { + "left": "pull_request:47280", + "right": "pull_request:47443", + "accept": false, + "reason": "MiniMax-M3-VL test fix is unrelated to the DeepSeek code cleanup." + }, + { + "left": "pull_request:47443", + "right": "pull_request:47444", + "accept": true, + "reason": "They share the same DeepSeek model-file cleanup and look like parallel variants of the same code-path fix." + }, + { + "left": "pull_request:47443", + "right": "pull_request:47445", + "accept": true, + "reason": "Same shared DeepSeek model-file change; the PEFT-related wrapper is a follow-up around the same fix." + }, + { + "left": "pull_request:47444", + "right": "pull_request:47445", + "accept": true, + "reason": "Both carry the same DeepSeek model-code cleanup and could plausibly be combined." + }, + { + "left": "pull_request:47291", + "right": "pull_request:47322", + "accept": false, + "reason": "Deprecating continuous batching config and Quark config-class changes are separate quantization issues." + }, + { + "left": "pull_request:47301", + "right": "pull_request:47304", + "accept": false, + "reason": "MTP generation docs and MPS graph-cache docs are unrelated documentation work." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47243", + "accept": false, + "reason": "OpenVINO exporter work and ExecuTorch exporter fixes are different exporter implementations." + }, + { + "left": "pull_request:47216", + "right": "pull_request:47301", + "accept": false, + "reason": "Compressed-tensors fp8 and MTP docs are unrelated changes." + }, + { + "left": "pull_request:47077", + "right": "pull_request:47304", + "accept": false, + "reason": "Japanese translation edits and MPS graph-cache docs are not the same artifact." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47143", + "accept": false, + "reason": "Nemotron Omni support and NeuCodec support are separate model additions." + }, + { + "left": "pull_request:46836", + "right": "pull_request:47387", + "accept": false, + "reason": "Unlimited OCR support and one-to-many conversion loading are unrelated." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47166", + "accept": false, + "reason": "Paged-attention XPU tests and quantization test extensions are different issues." + }, + { + "left": "pull_request:47210", + "right": "pull_request:47301", + "accept": false, + "reason": "Packed SDPA routing and MTP docs are unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47336", + "accept": false, + "reason": "Exporter feature work and offloading save bugfix are different code paths." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 1, + "component_count": 31, + "review_unit_index": 5, + "review_unit_count": 5, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-97", + "nodes": [ + "pull_request:40546", + "pull_request:43451", + "pull_request:45157", + "pull_request:45273", + "pull_request:46509", + "pull_request:46658", + "pull_request:46912", + "pull_request:47003", + "pull_request:47096", + "pull_request:47143", + "pull_request:47163", + "pull_request:47175", + "pull_request:47210", + "pull_request:47292", + "pull_request:47294", + "pull_request:47298", + "pull_request:47301", + "pull_request:47304", + "pull_request:47306", + "pull_request:47322", + "pull_request:47367", + "pull_request:47387", + "pull_request:47452" + ], + "soft_pairs": [ + "pull_request:47304|pull_request:47367", + "pull_request:46509|pull_request:47387", + "pull_request:46912|pull_request:47298", + "pull_request:46509|pull_request:47301", + "pull_request:47175|pull_request:47298", + "pull_request:47163|pull_request:47210", + "pull_request:45157|pull_request:47387", + "pull_request:47294|pull_request:47298", + "pull_request:46658|pull_request:47387", + "pull_request:47143|pull_request:47163", + "pull_request:46658|pull_request:47143", + "pull_request:46658|pull_request:47322", + "pull_request:47143|pull_request:47387", + "pull_request:47143|pull_request:47322", + "pull_request:45157|pull_request:47143", + "pull_request:47301|pull_request:47367", + "pull_request:47163|pull_request:47292", + "pull_request:46912|pull_request:47452", + "pull_request:47306|pull_request:47367", + "pull_request:40546|pull_request:43451", + "pull_request:43451|pull_request:47143", + "pull_request:43451|pull_request:47163", + "pull_request:43451|pull_request:46509", + "pull_request:45273|pull_request:47298", + "pull_request:43451|pull_request:47096", + "pull_request:43451|pull_request:47003", + "pull_request:47003|pull_request:47096", + "pull_request:43451|pull_request:47322", + "pull_request:43451|pull_request:47306", + "pull_request:47096|pull_request:47306", + "pull_request:47003|pull_request:47306", + "pull_request:47096|pull_request:47387", + "pull_request:47003|pull_request:47387", + "pull_request:47096|pull_request:47143", + "pull_request:47096|pull_request:47163" + ], + "prepared_review_unit_hash": "0c0d47f19b70de9deb41ab386142f993ac7ca9ecd0a271f2b1282bd7848129f2", + "component_budget": { + "node_count": 97, + "item_count": 97, + "soft_pair_count": 277, + "serialized_chars": 338403, + "estimated_input_tokens": 84601, + "estimated_eval_tokens": 169458 + }, + "budget": { + "node_count": 23, + "item_count": 23, + "soft_pair_count": 35, + "serialized_chars": 64456, + "estimated_input_tokens": 16114, + "estimated_eval_tokens": 32484 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a loose semantic grouping of unrelated PRs: new model integrations, exporter/runtime work, docs-only changes, and test/CI fixes. None of the soft edges look like the same concrete bug or change, so they should all remain separate.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "There is no true canonical PR for this cluster because the items are unrelated; if a representative has to be named, 43451 ('Add Molmo2') is the most complete self-contained feature PR (docs + tests + model plumbing), but it does not subsume the others.", + "best_issue_reason": null, + "best_pr_reason": "43451 is the best representative anchor only in the weak sense that it is a substantial, self-contained implementation PR. It is not a duplicate of the other items, but it is the clearest standalone feature PR in the set.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47304", + "right": "pull_request:47367", + "accept": false, + "reason": "Both are docs-related, but one is about MPS graph cache docs and the other changes pipeline device placement semantics; different changes." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47387", + "accept": false, + "reason": "Nemotron Omni model support vs one-to-many conversion infrastructure; different layers and different code paths." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47298", + "accept": false, + "reason": "RoPE test expansion across models vs MI450 failure fixes; only a broad testing theme, not the same change." + }, + { + "left": "pull_request:46509", + "right": "pull_request:47301", + "accept": false, + "reason": "Nemotron Omni support and MTP docs are unrelated feature areas." + }, + { + "left": "pull_request:47175", + "right": "pull_request:47298", + "accept": false, + "reason": "Olmo3 CI expectation updates vs MI450 model/runtime fixes; not the same underlying bug or change." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47210", + "accept": false, + "reason": "Both concern attention/XPU, but 47163 is test coverage and 47210 is runtime routing to varlen attention; different concrete changes." + }, + { + "left": "pull_request:45157", + "right": "pull_request:47387", + "accept": false, + "reason": "Prism GGUF dequantization support vs generic one-to-many conversion support; unrelated." + }, + { + "left": "pull_request:47294", + "right": "pull_request:47298", + "accept": false, + "reason": "Native causal_conv1d kernel plumbing vs MI450 model failures; different problems." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47387", + "accept": false, + "reason": "Step3.7 checkpoint conversion tooling vs core one-to-many conversion support; related only at a very high level." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47163", + "accept": false, + "reason": "NeuCodec model support vs paged-attention XPU tests; different features." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47143", + "accept": false, + "reason": "Step3.7 flash support and NeuCodec support are separate model integrations." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47322", + "accept": false, + "reason": "Step3.7 flash conversion/model work vs Quark config-class changes; unrelated." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47387", + "accept": false, + "reason": "NeuCodec model support vs one-to-many conversion infra; not the same change." + }, + { + "left": "pull_request:47143", + "right": "pull_request:47322", + "accept": false, + "reason": "NeuCodec model support vs AMD Quark config changes; unrelated." + }, + { + "left": "pull_request:45157", + "right": "pull_request:47143", + "accept": false, + "reason": "Prism GGUF dequantization and NeuCodec model support touch different subsystems." + }, + { + "left": "pull_request:47301", + "right": "pull_request:47367", + "accept": false, + "reason": "MTP docs vs pipeline device-placement clarification; different documentation topics and code semantics." + }, + { + "left": "pull_request:47163", + "right": "pull_request:47292", + "accept": false, + "reason": "Both are XPU test-related, but they target different models/features and different failures." + }, + { + "left": "pull_request:46912", + "right": "pull_request:47452", + "accept": false, + "reason": "RoPE tests vs broad Mamba standardization/test coverage; similar testing theme only, not the same issue." + }, + { + "left": "pull_request:47306", + "right": "pull_request:47367", + "accept": false, + "reason": "Neuron experimentation scripts vs pipeline device placement; unrelated." + }, + { + "left": "pull_request:40546", + "right": "pull_request:43451", + "accept": false, + "reason": "VibeVoice and Molmo2 are different model additions with separate docs, configs, and code." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47143", + "accept": false, + "reason": "Molmo2 and NeuCodec are distinct model integrations." + }, + { + "left": "pull_request:43451", + "right": "pull_request:46509", + "accept": false, + "reason": "Molmo2 and Nemotron Omni are unrelated model-support PRs." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47096", + "accept": false, + "reason": "Molmo2 model integration vs Kimi exportability work; different concrete changes." + }, + { + "left": "pull_request:45273", + "right": "pull_request:47298", + "accept": false, + "reason": "Liger eval OOM fix in trainer vs MI450 model failures; different runtime issues." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47003", + "accept": false, + "reason": "Molmo2 model support vs OpenVINO exporter support; different code paths and goals." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47096", + "accept": false, + "reason": "OpenVINO exporter addition and exportability fixes for Kimi are related to export, but they are not the same concrete change." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47306", + "accept": false, + "reason": "OpenVINO exporter work vs Neuron experiments; unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47387", + "accept": false, + "reason": "OpenVINO exporter support vs one-to-many conversion support; different infrastructure changes." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47306", + "accept": false, + "reason": "Kimi exportability work vs Neuron experiments; unrelated." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47387", + "accept": false, + "reason": "Kimi exportability vs one-to-many conversion infrastructure; not the same code-path problem." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47143", + "accept": false, + "reason": "Kimi exportability work vs NeuCodec support; different model/feature areas." + }, + { + "left": "pull_request:47096", + "right": "pull_request:47163", + "accept": false, + "reason": "Exportability work vs paged-attention XPU tests; unrelated." + }, + { + "left": "pull_request:47003", + "right": "pull_request:47322", + "accept": false, + "reason": "OpenVINO exporter support vs AMD Quark config changes; unrelated." + }, + { + "left": "pull_request:43451", + "right": "pull_request:47322", + "accept": false, + "reason": "Molmo2 model integration vs Quark config-class changes; unrelated." + }, + { + "left": "pull_request:46658", + "right": "pull_request:47143", + "accept": false, + "reason": "Step3.7 flash conversion tooling and NeuCodec support are different model projects." + } + ] + }, + "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": 31, + "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": "bc756db4e3e48839b65ec0e06cea7840a1f8d4b34ba32f8b8239c1d4da0dfe69", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4675, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated feature PRs: one adds covariate forecasting to TimesFM, the other adds DINOv3 image-classification support. The soft similarity is superficial, so this is not a duplicate pair.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 40755 is the stronger representative of this cluster by diff scope and implementation depth; PR 41224 is a separate DINOv3 feature PR, not the same change.", + "best_issue_reason": null, + "best_pr_reason": "PR 40755 is the best PR candidate here because it has the broader, more self-contained implementation across model code, utilities, tests, and docs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Reject: these touch different models and code paths (TimesFM covariates forecasting vs DINOv3 image classification support) and could not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 3, + "component_count": 31, + "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": 31, + "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": "8e68dd5f635a158e23994eefb2fa240f800e5720cab57a2eccecd381910c6c17", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5935, + "estimated_input_tokens": 1484, + "estimated_eval_tokens": 3224 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related GitHub Actions dependency bumps, but they address different actions and different failure surfaces: Node 24 compatibility for core workflow actions vs. docker/build-push-action version upgrades.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 42975 is the broader, more representative workflow-actions upgrade: it updates checkout/cache/upload-artifact across many workflows and matches the cluster\u2019s overall Node 24 compatibility theme.", + "best_issue_reason": null, + "best_pr_reason": "PR 42975 is the best representative PR because it covers the wider set of workflow-action incompatibilities and is more likely to serve as the umbrella change for this cluster.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "They are both workflow maintenance upgrades, but not the same concrete change: 42975 updates general GitHub Actions used for Node 24 compatibility, while 42976 only bumps docker/build-push-action. They could be merged independently and do not fix the same underlying bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 31, + "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": "15beb60189fbedc3256ed1ec20bf0c63457d374f62a97fec4b56a3faf3a5ca92", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13043, + "estimated_input_tokens": 3261, + "estimated_eval_tokens": 6778 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 4, + "serialized_chars": 13043, + "estimated_input_tokens": 3261, + "estimated_eval_tokens": 6778 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All four items are independent model-addition pull requests with overlapping Transformers boilerplate, not duplicate fixes or the same concrete change. The shared files are mostly docs, auto-mapping, and repo-check registration, which is expected for new-model integrations.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "ViT NEPA and Molmo are unrelated model additions with different model families, files, and implementation details. They only overlap on generic docs/auto-registration paths, which is not enough to treat them as the same PR." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "These add different models in different domains (vision vs. time series) and touch distinct codepaths. Shared Transformers boilerplate does not indicate the same underlying change." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM are separate model integrations with different architectures and support files. Their common filenames reflect standard new-model wiring, not a shared bug or duplicate change." + }, + { + "left": "pull_request:45490", + "right": "pull_request:46562", + "accept": false, + "reason": "CTSM and Step3p7 are unrelated model additions. They share registration/documentation machinery, but the concrete model code, processing, and intended behavior are different, 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": 6, + "component_count": 31, + "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": "4969ddec6465eec7f8cac5a6531d9d3b9b1bd73ce01ecff971e748386ad0a7a1", + "component_budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19581, + "estimated_input_tokens": 4896, + "estimated_eval_tokens": 10048 + }, + "budget": { + "node_count": 7, + "item_count": 7, + "soft_pair_count": 15, + "serialized_chars": 19581, + "estimated_input_tokens": 4896, + "estimated_eval_tokens": 10048 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Almost all PRs are unrelated despite superficial test/subsystem overlap. The only clear duplicate pair is the two YOLOS device_map=\"auto\" fixes, which address the same device mismatch in the same model path.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46922 is the better canonical representative: it fixes the YOLOS device mismatch directly in forward() and also adds the no-split-module change, making it the more complete version of the same fix target.", + "best_issue_reason": null, + "best_pr_reason": "PR 46922 is the strongest overall candidate because it most directly addresses the YOLOS device_map auto failure and subsumes the narrower 46886 change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Different models and different changes; one is a Whisper tokenizer/test fix, the other is unrelated test edits for CLIP and ViT." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Whisper decoding logic is unrelated to the FP8 dequant/loading changes in the other PR." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Both mention Whisper-related code, but one is a tokenizer decode fix while the other is generation/watermarking validation and broader model changes; not the same bug." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46922", + "accept": false, + "reason": "Whisper tokenizer/tests and YOLOS device placement are unrelated issues." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46539", + "accept": false, + "reason": "Both are test/CI-oriented, but they target different failures in different models and code paths." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "Unrelated test assertions versus FP8 dequantization/loading internals." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "These touch different subsystems and different failure modes; the overlap is only that both include test updates." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46922", + "accept": false, + "reason": "CLIP/ViT test edits are unrelated to the YOLOS device mismatch fix." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46539", + "accept": false, + "reason": "FP8 kernel/loading fixes and Phimoe CI memory patch address different problems." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "Different code paths: FP8 dequantization versus generation config/watermarking and related model fixes." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46922", + "accept": false, + "reason": "No concrete overlap beyond both being model-related fixes; the bugs and code paths are different." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46539", + "accept": false, + "reason": "Generation/watermarking validation is unrelated to the Phimoe CI memory workaround." + }, + { + "left": "pull_request:46380", + "right": "pull_request:46922", + "accept": false, + "reason": "Different models and failure modes; watermarking/config validation does not match the YOLOS device placement bug." + }, + { + "left": "pull_request:46539", + "right": "pull_request:46922", + "accept": false, + "reason": "CI runner memory limits for Phimoe are unrelated to YOLOS device_map auto behavior." + }, + { + "left": "pull_request:46886", + "right": "pull_request:46922", + "accept": true, + "reason": "Both PRs fix the same YOLOS device_map auto device-mismatch problem in the same model file; 46922 simply includes a more complete version of the same fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted pair (46886 vs 46922) is a plausible duplicate/superset relationship for the same YOLOS device_map='auto' bugfix. The other soft-edge verdicts are conservative and do not overclaim mergeability or shared bug identity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 7, + "component_count": 31, + "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": "ed0126c680022c1a455e6eb01db4f17b7296a91426018407efa260e57645cf5e", + "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 are separate \"[don't merge]\" test/CI PRs, but they change different workflows or setup code paths. The soft similarities are superficial; none looks like the same underlying bug or change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No PR is a true canonical duplicate here; the cluster is a false-positive grouping of unrelated test/CI changes. If one must be picked as the most central, #43532 is the broadest workflow-centric change, but it is still not a duplicate of the others.", + "best_issue_reason": null, + "best_pr_reason": "#43532 is the closest thing to a hub because it spans several CI/testing files, but its actual change set is still distinct from the setup/version tweak in #43488 and the security-gate workflow change in #46747.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:46747", + "accept": false, + "reason": "Different areas and purposes: #43488 changes setup/dependency/version/test scaffolding, while #46747 only flips the security-gate workflow trigger. No shared code path or mergeable common fix." + }, + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "The only overlap is a shared test file, but the actual edits are unrelated: #43488 updates setup/dependencies/version, while #43532 rewrites CI/workflow logic and related utilities. Not the same change." + }, + { + "left": "pull_request:43532", + "right": "pull_request:46747", + "accept": false, + "reason": "Both touch GitHub Actions, but they modify different workflows with different behavior: self-comment/comment-driven CI versus the security-gate trigger. Too different to be 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": 8, + "component_count": 31, + "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": "fe65d6731ae2fb18789487ecfa3ea19d5ff38809231ab25dd3d39afc5d7711a2", + "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 related but distinct feature PRs: one adds Promptable Concept Segmentation, the other adds Promptable Visual Segmentation. They share framework/docs scaffolding, but the actual pipelines, model hooks, and tests are for different code paths.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No canonical PR: the two PRs implement different segmentation features and do not describe the same underlying change.", + "best_issue_reason": null, + "best_pr_reason": "PR 43613 is the broader standalone addition, but it is not a duplicate of 43612; both should be tracked separately.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Different features and different pipeline modules/tests; shared docs/registry edits are incidental and do not indicate the same concrete change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 9, + "component_count": 31, + "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": "a2cce220e38e2991ac90bdcac6ab5fa5b5c2f9600dd08f746d4b05fdf8c4c8a3", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4292, + "estimated_input_tokens": 1073, + "estimated_eval_tokens": 2402 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs in the same file, but they address different Trainer extension points: loss scaling for gradient accumulation vs. signature-column filtering for custom/PEFT trainers. They are related at a subsystem level, not the same underlying change.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Both patch trainer extensibility in src/transformers/trainer.py, but one fixes gradient-accumulation loss scaling and adds _loss_is_scaled_for_ga, while the other extracts _get_signature_columns for input filtering. Different code paths and different behaviors, so not 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": 10, + "component_count": 31, + "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": "a329184fe5d044e39c1fd5eacfca763360fd9950165c48cc8121aa4d385794bc", + "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 touch `src/transformers/cli/serve.py`, but they fix different `TypeAdapter`-related problems and should not be deduped.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43842 is the more concrete and targeted fix: it explicitly guards the `TypeAdapter` import behind `TYPE_CHECKING` and quotes the annotation to avoid a runtime NameError when pydantic is absent.", + "best_issue_reason": null, + "best_pr_reason": "PR #43842 is the best representative of the cluster because it addresses a specific import-time failure mode with a narrowly scoped fix; PR #43656 is a broader typing cleanup that appears to target a different NameError path.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Same file and similar symptom, but the underlying bugs differ: #43656 is about general typing/annotation NameError cleanup, while #43842 specifically fixes `TypeAdapter` import handling when pydantic is not installed. They are not the same change and would not plausibly merge into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 11, + "component_count": 31, + "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": 12, + "component_count": 31, + "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:44019|pull_request:44085", + "pull_request:44086|pull_request:44722", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44018|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44068|pull_request:44086", + "pull_request:44085|pull_request:44086", + "pull_request:44059|pull_request:44086", + "pull_request:44059|pull_request:44722", + "pull_request:44098|pull_request:44101", + "pull_request:44018|pull_request:44071", + "pull_request:44017|pull_request:44073", + "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44722", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44017|pull_request:44071", + "pull_request:44071|pull_request:44086", + "pull_request:44073|pull_request:44101", + "pull_request:44073|pull_request:44161", + "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:43997|pull_request:44017", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44018", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request:44018|pull_request:44161", + "pull_request: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" + ], + "prepared_review_unit_hash": "950019d22b1e6489cd4472b85ce802ce62e326d8da485532cd4fe24bc27b1240", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84514, + "estimated_input_tokens": 21129, + "estimated_eval_tokens": 42514 + }, + "budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 72, + "serialized_chars": 76945, + "estimated_input_tokens": 19237, + "estimated_eval_tokens": 38730 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a loose theme of output-tracing refactors across many unrelated model families. The only clear duplicate groups are the GPT-J/CodeGen trio and the GPT-Neo pair; the rest share a tracking issue or broad subsystem but not the same concrete code change.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "44066 is the clearest representative of the GPT-J/CodeGen refactor: explicit issue linkage, concrete code-path changes in both files, and a focused title matching the actual diff.", + "best_issue_reason": null, + "best_pr_reason": "44066 is the strongest single representative in the cluster because it is the most directly scoped, issue-linked, and concrete implementation of the standardized output-tracing change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model families (CVT/FNet vs GPT-J/CodeGen); shared issue target alone is not a duplicate." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "CVT/FNet output-tracing refactor is unrelated to DeBERTa's attention/output handling." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing refactor on the same concrete code path; plausible to merge as one PR." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "Both are vision backbones, but they change different models and code paths; too broad to treat as duplicates." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet standardized outputs and GPT-J/CodeGen tracing are unrelated code changes." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen refactor vs EfficientNet refactor; same theme, different concrete change." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are unrelated implementations; only the issue link is common." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44071", + "accept": false, + "reason": "MGP-STR and MPT touch different model code paths, so they are not the same change." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen output tracing and MPT output tracing are separate model-specific refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Both are GPT-Neo output-tracing refactors in the same file and same code path." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen vs GPT-Neo; same abstraction, different model implementation." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet output tracing and GPT-J/CodeGen output tracing are not the same underlying change." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J/CodeGen are unrelated model-specific refactors." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are unrelated implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are different concrete code paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet output tracing and MPT output tracing are separate refactors in different models." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert are unrelated model changes." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert do not share the same code path." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are different model families and different implementations." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44722", + "accept": false, + "reason": "MPT output tracing and GPT-J/CodeGen output tracing are unrelated changes." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44085", + "accept": false, + "reason": "GPT-Neo vs GPT-J/CodeGen; similar refactor pattern, but not the same PR." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated model-specific changes." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44085", + "accept": false, + "reason": "MPT and GPT-J/CodeGen differ in model and code path." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are different concrete implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-2 output tracing and MGP-STR output tracing are separate model changes." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-2 and GPT-J/CodeGen are different code paths; same refactor style is not enough." + }, + { + "left": "pull_request:44098", + "right": "pull_request:44101", + "accept": false, + "reason": "ViLT and Flaubert are unrelated model implementations." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44073", + "accept": false, + "reason": "SegFormer and VisualBert are different model code paths." + }, + { + "left": "pull_request:44017", + "right": "pull_request:44161", + "accept": false, + "reason": "SegFormer and LongT5 are unrelated changes despite both using tracing decorators." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-2 and MPT output tracing are different concrete implementations." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44101", + "accept": false, + "reason": "MPNet and Flaubert are unrelated model-specific refactors." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44076", + "accept": false, + "reason": "MPT and ImageGPT are different model families and different code paths." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert are unrelated model changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are separate concrete changes." + }, + { + "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 code paths." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44161", + "accept": false, + "reason": "TextNet and LongT5 are unrelated model-specific refactors." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are different models with different output-tracing logic." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are unrelated refactors." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert are different concrete changes." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are not the same underlying bug/change." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT touch different code paths." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are unrelated model implementations." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBert are different model-specific refactors." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44044", + "accept": false, + "reason": "GPT-Neo and DeBERTa are unrelated code changes." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are separate output-tracing refactors." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen standardized tracing change; the diffs target the same code path and could be merged into one PR." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both appear to be the GPT-J/CodeGen output-tracing refactor on the same files and same underlying change." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J/CodeGen and SpeechT5 are unrelated model-specific changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "CVT/FNet and ResNet are different model families; only the thematic issue is shared." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet output-tracing refactor and SpeechT5 output-tracing refactor are separate changes." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa and SpeechT5 are unrelated concrete code paths." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa and Vits are different model implementations." + }, + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "FocalNet and ImageGPT are unrelated model-specific refactors; same decorator pattern is not enough." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary stays within the packet evidence: it identifies the strongest duplicate groupings conservatively and avoids claiming unrelated model refactors are duplicates. The accepted soft-edge pairs all look mergeable as the same output-tracing bugfix within the same model/file scope." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 12, + "component_count": 31, + "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:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "a2541232114ce2c148d69200a43337730c2cd41f2805be6d604d06a4e4070565", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 77, + "serialized_chars": 84514, + "estimated_input_tokens": 21129, + "estimated_eval_tokens": 42514 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 5, + "serialized_chars": 13069, + "estimated_input_tokens": 3268, + "estimated_eval_tokens": 6792 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All six PRs are model-specific refactors around standardized output tracing/capture_outputs, but they touch different architectures and code paths. They are related thematically, not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: each change targets a different model implementation (RegNet, SqueezeBert, MPNet, VisualBert, TextNet, SpeechT5) with separate output-tracing mechanics.", + "best_issue_reason": null, + "best_pr_reason": "44129 is the strongest standalone representative because it has an explicit issue target and a clearly scoped SpeechT5 refactor, but it is still not a duplicate of the other model-specific PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "Same refactor theme, but VisualBert and TextNet are different models with different forward/output paths; not the same underlying change." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet vs VisualBert: distinct architectures and output-tracing implementations, so they should not be merged as one PR." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "MPNet and TextNet are separate model code paths; shared capture_outputs usage is too broad to treat as duplicate work." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert are unrelated model implementations; this is a common refactor pattern, not the same bug/change." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 fix different model-specific output tracing behavior; the explicit issue link on 44129 does not make them 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": 13, + "component_count": 31, + "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": "84a96b45dad6c1ef4d356fb353c60feb0df670c9b92ffc157dd96f4e5895db5f", + "component_budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40180, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 + }, + "budget": { + "node_count": 13, + "item_count": 13, + "soft_pair_count": 40, + "serialized_chars": 40180, + "estimated_input_tokens": 10045, + "estimated_eval_tokens": 20346 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All items are model-specific output-tracing refactors. They share a common pattern and helper usage, but they target different architectures and code paths, so they are not true duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44003 is the broadest representative here: it spans both `falcon_mamba` and `mamba` and adds shared output-capture plumbing, so it best anchors the cluster.", + "best_issue_reason": null, + "best_pr_reason": "44003 is the most central and infrastructure-like PR in the set, making it the best single representative even though the others are distinct changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families (`timm_backbone` vs `upernet`) and different forward paths; same tracing theme only." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "`timm_backbone` and `vision_text_dual_encoder` are unrelated code paths; not the same bug/change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different architectures (`timm_backbone` vs `univnet`); only the refactor pattern is similar." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "`timm_backbone` refactor vs `mamba/falcon_mamba` output tracing; separate model implementations and changes." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "`timm_backbone` and `rwkv` are distinct model families with different concrete forward paths." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "`mobilenet_v1` and `univnet` are unrelated models; same refactor motif, not the same change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "`mobilenet_v1` vs `upernet` touches different architectures and output code paths." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "`mobilenet_v1` and `depth_anything` are different model families; not mergeable as one duplicate fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "`mobilenet_v1` refactor is unrelated to the `vision_encoder_decoder` wrapper change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "`mobilenet_v1` and `speech_encoder_decoder` are separate code paths; only the refactor style overlaps." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "`vision_text_dual_encoder` and `univnet` are different model families with different forward logic." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "`vision_text_dual_encoder` vs `upernet` is a different model-specific refactor, not a duplicate." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different concrete implementations (`vision_text_dual_encoder` vs `mamba/falcon_mamba`), so not the same change." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "`vision_text_dual_encoder` and `depth_anything` are unrelated models; same tracing theme only." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "`univnet` and `upernet` are distinct architectures; not the same underlying bug or fix." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "`univnet` vs `mamba/falcon_mamba` affects separate code paths and files." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "`univnet` and `depth_anything` are unrelated model-specific changes." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "`upernet` and `rwkv` are different models with separate output handling paths." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "`upernet` refactor is unrelated to the `mamba/falcon_mamba` output-tracing change." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "`mamba/falcon_mamba` vs `superpoint` are different architectures and different concrete code paths." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "`mamba` refactor and `codegen` refactor are separate model changes; not a single duplicate fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "`timm_backbone` and `codegen` are unrelated implementations; same helper usage does not make them duplicates." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "`univnet` and `codegen` touch different code paths and model families." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "`upernet` and `codegen` are separate model-specific refactors." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "`vision_text_dual_encoder` and `codegen` are unrelated code paths; not the same underlying change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "`depth_anything`/`prompt_depth_anything` vs `vision_encoder_decoder` are different model families and forward paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "`depth_anything`/`prompt_depth_anything` is unrelated to `speech_encoder_decoder`; same refactor pattern only." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "`depth_anything` and `superpoint` are separate vision models with different output shapes and code paths." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "`depth_anything` refactor is unrelated to `rwkv`; different model family and implementation." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "`depth_anything` and `dpr` are different model families; not a duplicate change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "`vision_encoder_decoder` and `rwkv` have no shared concrete bug/change; only thematic similarity." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Both are encoder-decoder wrappers, but vision and speech wrappers are separate concrete code paths and the diffs are model-specific." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "`vision_encoder_decoder` and `superpoint` are unrelated model implementations." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "`vision_encoder_decoder` and `dpr` are different model families and output paths." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "`speech_encoder_decoder` and `superpoint` are different architectures; not the same duplicate fix." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "`speech_encoder_decoder` and `rwkv` are unrelated changes in different model families." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "`speech_encoder_decoder` and `dpr` are separate code paths; same helper refactor does not make them duplicates." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "`superpoint` and `rwkv` are distinct models with different forward logic." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "`superpoint` and `dpr` are different model families and not the same underlying issue." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "`rwkv` and `dpr` are unrelated model-specific refactors; no shared concrete fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 31, + "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": "1d098fdbe3f0899206ce04e6326eff5483b087774580635c78e1795909b8b348", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6551, + "estimated_input_tokens": 1638, + "estimated_eval_tokens": 3532 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated model-addition PRs that only overlap on standard Transformers registration/docs plumbing. They do not appear to fix or implement the same underlying change, so the soft similarity link should be rejected.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45401 is the clearer representative: it is scoped to a single model integration, has an explicit issue target, and includes the full implementation/test wiring for Voxtral TTS.", + "best_issue_reason": null, + "best_pr_reason": "PR 45401 is the stronger standalone PR for this cluster because it is more focused and traceable, whereas PR 44662 is a separate PenguinVL model addition with no concrete relation to Voxtral TTS.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different model integrations (PenguinVL vs Voxtral TTS) with distinct codepaths and purposes; the shared files are generic model-registration/docs changes, not evidence of 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": "pull_request", + "component_index": 15, + "component_count": 31, + "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": "7607ea219c1b6c691e71890acd34e29e3e51c9531efbd52eb116bf1026d89a56", + "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": "These are related fusion-feature PRs, but they are not the same change: one introduces a standalone Module Fusion API, while the other wires n-to-1 kernel fusion into `KernelConfig` and model loading. They overlap conceptually yet target different implementation paths and could not safely be merged as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45363 is the broader, more concrete implementation: it integrates kernel fusion through `KernelConfig`, model loading, and integration code rather than only adding a standalone API scaffold.", + "best_issue_reason": null, + "best_pr_reason": "PR #45363 is the best representative for this cluster because it covers the end-to-end fusion workflow and has the more actionable code-path change; PR #44979 is more of an isolated API foundation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Related feature area, but different scope and code paths: #44979 adds a new `module_fusion.py` API layer, while #45363 implements n-to-1 kernel fusion via `KernelConfig` and integration hooks. They are not the same concrete 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": 16, + "component_count": 31, + "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": "c29d237b5fbb5dad68de55612aa709de3af05dc728590f9b6db6c7b15cf13452", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests fix the same AttributeError in `_patch_mistral_regex` by switching from `tokenizer.backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; the later PR adds tests and is the cleaner representative.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the stronger canonical PR: it contains the same core fix as 45086 plus test coverage, making it the more complete representative of this change.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the best representative because it covers the full bug fix and validation, while 45086 is an earlier, smaller version of the same patch.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same underlying bug and identical code-path fix in `_patch_mistral_regex`; 45317 is just the more complete version with tests, so these are duplicate/near-duplicate PRs." + } + ] + }, + "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": 31, + "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": "507fdc7919b9fad262e07c8662eeb1336ab5971226252e39d314797ab48aad35", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6066, + "estimated_input_tokens": 1517, + "estimated_eval_tokens": 3290 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs touch Nemotron H, but they address different scopes: one is a narrow conversion-mapping/refactor change, the other is a broader model-implementation and docs split into dense/sparse variants. They are related but not duplicates.", + "confidence": 0.88, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45569 is the better canonical PR because it is the broader, more complete Nemotron H change: it introduces the dense/sparse model packages, updates auto-mapping, docs, and tests, whereas #45115 is a narrower structural conversion/refactor patch.", + "best_issue_reason": null, + "best_pr_reason": "PR #45569 best represents the cluster since it covers the main user-facing/modeling change and associated documentation/test updates; #45115 appears to be a supporting conversion-mapping refactor rather than the primary implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "They are not the same concrete code-path fix. #45115 adds checkpoint conversion key renames for NemotronH/Bamba compatibility, while #45569 implements the Nemotron H split into dense/sparse models with docs and tests. Overlap in files does not make them 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": 18, + "component_count": 31, + "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": "c3c69b77f4fd108d0ab2fc1d7cf9f0e91f0ec5ff69516b1ea95aee85ea387fa5", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5514, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are both typing-related PRs in transformers, but they are not the same concrete change. PR 45415 is a broad type-checking sweep plus pyproject ignore changes across many files, while PR 45425 adds specific typing helpers/protocols and targeted integration fixes. I would not merge them as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "45425 is the more coherent, code-focused typing change set; 45415 is a much broader mechanical type-checking sweep with config noise across many files.", + "best_issue_reason": null, + "best_pr_reason": "45425 is the better representative because it targets a specific typing problem space with tighter scope and clearer code-path intent.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Both are about typing, but they fix different things: 45415 is a repository-wide type-checking cleanup, while 45425 introduces typing protocol aliases and specific model/integration adjustments. Shared filenames alone are not enough to treat them as the same PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 31, + "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": "47c14f91fe721fafb8cf69f302956f4275635313a0cd7b6c493d497a15a931de", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4507, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4507, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are EP-related, but they fix different bugs: one addresses EP/DeepSpeed ZeRO-3 loading and config serialization, while the other fixes expert tensors being overwritten during FSDP2 rank-0 broadcast. Shared files alone aren\u2019t enough to treat them as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "pull_request:45662 is the better representative because it targets a concrete, well-scoped correctness bug (experts being overwritten by rank-0 broadcast) with a clearer end-to-end failure mode and broader EP/FSDP2 handling changes.", + "best_issue_reason": null, + "best_pr_reason": "pull_request:45662 is the strongest standalone PR in this cluster: it fixes a specific data corruption issue in distributed EP loading, touches the relevant loading and MoE/tensor-parallel paths, and is more clearly a single mergeable change than the other PR.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "They share the EP/distributed-loading theme, but the underlying bugs differ: 45548 is about EP + DeepSpeed ZeRO-3 loading and stripping distributed_config from serialization, while 45662 fixes experts being silently overwritten by rank-0 broadcast under FSDP2. Not the same code-path problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 20, + "component_count": 31, + "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": "df5ebd1a64c5b46f7663b64b4779686c7300f32a1fdba068fa01687084ead0b1", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4673, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4673, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs target the same xdist collision around `captured_info` debug artifacts in the testing utilities. PR 45645 is the broader follow-up/superset: it includes the core file-name fix plus CI workflow updates to collect the new per-worker logs, while 45639 is the narrower core fix alone.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the best canonical representative because it contains the underlying xdist-safe logging fix and the necessary CI artifact handling changes, making it the more complete version of the change.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 best captures the full user-facing change set: fixing the collision in `testing_utils.py` and updating workflow/log collection so the new files are actually preserved in CI.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Both address the same concrete bug: `captured_info` writes colliding under xdist. The second PR is a superset that adds workflow support for the same per-worker log files, so they could plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 31, + "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": 22, + "component_count": 31, + "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": "d2c60d37ab43a6afb03b2a42156dad5512192a9239d16de2abb67386103645b0", + "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": "These PRs are related distributed-work items, but they are not the same change: #46269 is a broad distributed branch base, while #46126 is a narrower EP + Trainer integration built on top of that foundation. The shared filenames look like branch overlap rather than duplicate intent.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "#46269 is the broader umbrella PR introducing the distributed branch base and multiple infra changes, so it is the better canonical representative of the cluster.", + "best_issue_reason": null, + "best_pr_reason": "#46269 is the best representative because it subsumes the shared distributed groundwork; #46126 is a specific follow-on feature PR, not a duplicate of the same concrete fix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Different scope and intent: #46126 adds EP/Trainer integration details, while #46269 establishes the distributed branch base. Shared files alone are not enough to treat them as the same PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 23, + "component_count": 31, + "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": "4877a81918c0d5ff80442ad8c639c66609c5acf868eee8e3bbc7232881f4bfa3", + "component_budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26664, + "estimated_input_tokens": 6666, + "estimated_eval_tokens": 13588 + }, + "budget": { + "node_count": 10, + "item_count": 10, + "soft_pair_count": 31, + "serialized_chars": 26664, + "estimated_input_tokens": 6666, + "estimated_eval_tokens": 13588 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This is a heterogeneous cluster of workflow-maintenance pull requests. The only apparent subcluster is pr-ci-security-gate.yml, but those PRs make different security changes and are not duplicate fixes. I would not merge any of the soft edges.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "Best representative is PR 46554: it makes the most substantive functional change to the shared PR CI security gate, while the rest are either unrelated workflow tweaks or narrower follow-up edits to different concerns.", + "best_issue_reason": null, + "best_pr_reason": "PR 46554 is the strongest standalone representative because it changes the core security-gate logic rather than a peripheral workflow setting.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46138", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflows and different problems: self-comment CI stdin handling vs PR CI security-gate hardening." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46558", + "accept": false, + "reason": "Unrelated files and fixes; one updates self-comment CI, the other adjusts PR CI gate permissions/flow." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46563", + "accept": false, + "reason": "Different workflow files and different changes; not the same underlying bug or change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46716", + "accept": false, + "reason": "Self-comment CI command invocation is unrelated to the PR CI security gate secrets change." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflows and behavior; one pipes comments to a helper script, the other changes CI image tagging logic." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "SSH runner hardening is unrelated to benchmark workflow trigger changes." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "Different workflows and different fixes; SSH runner secret handling vs anti-slop permissions scoping." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Unrelated workflow files; SSH runner secret cleanup is not the same change as reviewer assignment checkout ref selection." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46554", + "accept": false, + "reason": "Both touch CI security themes, but they fix different workflows and different code paths." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflow files and different security concerns; not duplicate changes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Anti-slop permissions scoping and benchmark dispatch support are unrelated workflow changes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Different workflow jobs with different behavior; not the same underlying change." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "Anti-slop workflow permissions and build-ci image selection are separate concerns." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46554", + "accept": false, + "reason": "Same broad CI area, but different workflows and different fixes." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Reviewer assignment checkout ref and benchmark dispatch support are unrelated." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Different workflows and different behavior; not duplicate." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflow files and different problems." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark trigger expansion and build-ci commit-message parsing are unrelated changes." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46554", + "accept": false, + "reason": "Different workflows and different code paths; no duplicate bug/change here." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46554", + "accept": false, + "reason": "Build-ci image tagging logic and PR CI security-gate logic are separate changes." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46558", + "accept": false, + "reason": "Different workflows and different fixes; not the same underlying issue." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46558", + "accept": false, + "reason": "Same file, but different security hardening edits in different sections; they are related, not duplicates." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46563", + "accept": false, + "reason": "Both are PR CI gate edits, but one adds checks while the other only annotates the reusable workflow; not the same change." + }, + { + "left": "pull_request:46554", + "right": "pull_request:46716", + "accept": false, + "reason": "Different modifications to the same workflow file; one adds gate logic, the other changes secret inheritance." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46563", + "accept": false, + "reason": "Same workflow file but different concerns: permissions/approval gating vs reusable-workflow commentary." + }, + { + "left": "pull_request:46558", + "right": "pull_request:46716", + "accept": false, + "reason": "Different security-gate edits in the same file, but not the same underlying fix." + }, + { + "left": "pull_request:46563", + "right": "pull_request:46716", + "accept": false, + "reason": "One adds safety notes, the other changes secrets handling; not duplicate." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46563", + "accept": false, + "reason": "Different workflow files and different fixes; no shared underlying change." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46563", + "accept": false, + "reason": "SSH runner changes are unrelated to the reusable workflow note update in PR CI." + }, + { + "left": "pull_request:46504", + "right": "pull_request:46716", + "accept": false, + "reason": "Build-ci docker image workflow changes are unrelated to PR CI secrets handling." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46716", + "accept": false, + "reason": "Different workflows and different concerns; not a duplicate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 31, + "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": 25, + "component_count": 31, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46310-2", + "nodes": [ + "pull_request:46310", + "pull_request:47162" + ], + "soft_pairs": [ + "pull_request:46310|pull_request:47162" + ], + "prepared_review_unit_hash": "2eb7a3c7eccb3a26c3f3331517d0371cd384db684a9df6d600ac2c37458f9db0", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4883, + "estimated_input_tokens": 1221, + "estimated_eval_tokens": 2698 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4883, + "estimated_input_tokens": 1221, + "estimated_eval_tokens": 2698 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs fix the same off-by-one in MoonshineStreaming\u2019s sliding_window_mask_function right window (`<` to `<=`) and target the same issue, with 47162 also adding a test and an extra same-pattern fix in Gemma4.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "47162 is the better canonical PR because it contains the same MoonshineStreaming fix as 46310, adds test coverage, and appears to be the more complete follow-up.", + "best_issue_reason": null, + "best_pr_reason": "47162 is the strongest representative of the cluster: it fixes the same MoonshineStreaming bug, includes validation, and is the more comprehensive change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46310", + "right": "pull_request:47162", + "accept": true, + "reason": "Same concrete off-by-one bug in MoonshineStreaming sliding_window_mask_function, same target issue, and the diffs overlap on the exact code path; they could plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 26, + "component_count": 31, + "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": "858dd56a959c4004399061a39a41e1b2f455c432b60ad42527ad5f68c1dcf128", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4410, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4410, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two distinct tokenizer-related fixes: one corrects CodeLlama decode stripping behavior, the other fixes DeepSeek-Coder/DeepSeek-LLM loading under the wrong tokenizer class. They share a broad area but not the same bug or code path.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:46574", + "right": "pull_request:46580", + "accept": false, + "reason": "Different underlying problems and code paths: 46574 changes CodeLlama decoder setup to preserve leading spaces, while 46580 adds checkpoint-prefix routing in AutoTokenizer for specific DeepSeek models. 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": 27, + "component_count": 31, + "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-46903-3", + "nodes": [ + "pull_request:46903", + "pull_request:46958", + "pull_request:47090" + ], + "soft_pairs": [ + "pull_request:46903|pull_request:47090", + "pull_request:46903|pull_request:46958", + "pull_request:46958|pull_request:47090" + ], + "prepared_review_unit_hash": "d4bf563404603c5a7b0fa8d638e628c891936835c2fc59bf154bbf5428416778", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8781, + "estimated_input_tokens": 2196, + "estimated_eval_tokens": 4648 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8781, + "estimated_input_tokens": 2196, + "estimated_eval_tokens": 4648 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are related follow-ups around training-loss double-shift handling, but they do not read like one clean duplicate cluster: 46903 also covers PPFormulaNet, 46958 is a CohereASR-only CrossEntropyLoss rewrite, and 47090 is a broader later loss-plumbing change that adds Moonshine variants. Treat them as overlapping fixes rather than one clearly mergeable single PR set.", + "confidence": 0.63, + "canonical_issue_reason": null, + "canonical_pr_reason": "47090 is the strongest canonical PR because it covers the broadest affected code paths, keeps the loss plumbing aware of batch-aware reduction, and includes the most complete test coverage for the expanded model set.", + "best_issue_reason": null, + "best_pr_reason": "47090 is the best standalone PR to keep: it addresses the shared loss-path problem with the most complete scope and test coverage across CohereASR and the later related variants.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46903", + "right": "pull_request:47090", + "accept": false, + "reason": "Related loss-bug family, but not clearly the same mergeable patch: 46903 bundles CohereASR plus PPFormulaNet, while 47090 revises the CohereASR loss path again and expands to Moonshine variants." + }, + { + "left": "pull_request:46903", + "right": "pull_request:46958", + "accept": false, + "reason": "Both touch CohereASR double-shift handling, but 46903 is a broader multi-model fix and 46958 is a different CE-only implementation, so they do not look like one clearly mergeable PR." + }, + { + "left": "pull_request:46958", + "right": "pull_request:47090", + "accept": false, + "reason": "They are in the same bug family, but 47090 is a later broader loss-plumbing change with additional model coverage rather than the same concrete patch as 46958." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The packet supports a related family of double-shift fixes, but the summary overstates how confidently they are the same bug and mergeable as one PR set. PR 46903 is tied to a different issue target and covers CohereASR plus PPFormulaNet; PR 46958 is a separate CohereASR-only implementation using CrossEntropyLoss; PR 47090 is a later broader patch that changes the loss plumbing again and adds Moonshine variants. That looks more like overlapping follow-up fixes than one clearly mergeable single-bug PR cluster." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 28, + "component_count": 31, + "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-47140-2", + "nodes": [ + "pull_request:47140", + "pull_request:47303" + ], + "soft_pairs": [ + "pull_request:47140|pull_request:47303" + ], + "prepared_review_unit_hash": "62992d23e77e028081d13227993c179e1b427a613697cae38db27698c2d4073c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4409, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4409, + "estimated_input_tokens": 1103, + "estimated_eval_tokens": 2462 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the chat CLI\u2019s management URL construction in `src/transformers/cli/chat.py`, specifically ensuring management endpoints are called against the service root instead of a `/v1` model API base. PR 47303 is the broader, cleaner fix and is the better canonical representative.", + "confidence": 0.87, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 47303 is the broader and more direct fix: it centralizes service-root derivation for chat management calls and routes the management request there, matching the underlying URL-base bug more generally than the narrower `/v1` normalization in PR 47140.", + "best_issue_reason": null, + "best_pr_reason": "PR 47303 is the best standalone representative because it captures the core chat management routing problem in a more general way and is less tied to a single `/v1` edge case.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47140", + "right": "pull_request:47303", + "accept": true, + "reason": "Same underlying chat CLI management URL bug in the same file: both adjust how management calls are built so they hit the service root rather than the model `/v1` base. The second PR is a broader formulation of the same fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: both PRs touch `src/transformers/cli/chat.py` and `tests/cli/test_chat.py`, and both address chat management URL construction around the `/v1` base/service-root split. The soft-edge verdict is conservative enough given the shared files and closely related diffs, and PR 47303 is reasonably described as the broader representative. No obvious overstatement beyond the evidence shown." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 29, + "component_count": 31, + "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-47261-2", + "nodes": [ + "pull_request:47261", + "pull_request:47262" + ], + "soft_pairs": [ + "pull_request:47261|pull_request:47262" + ], + "prepared_review_unit_hash": "35e32d12c736e43f5cb4554ded5978eb283105e173ad90557aa70034e7c5fa58", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4419, + "estimated_input_tokens": 1105, + "estimated_eval_tokens": 2466 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4419, + "estimated_input_tokens": 1105, + "estimated_eval_tokens": 2466 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two pull requests target unrelated bugs in different parts of the library: training loss scaling vs. generation logits warping. They should not be deduplicated or merged together.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for deduplication: PR 47261 fixes gradient inflation with label smoothing during gradient accumulation, while PR 47262 adds/validates `min_tokens_to_keep` in `TopHLogitsWarper`.", + "best_issue_reason": null, + "best_pr_reason": "PR 47261 is the stronger representative of its own bug report because it has an explicit issue link and directly addresses a concrete training-correctness defect; PR 47262 is a separate generation feature/bugfix.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47261", + "right": "pull_request:47262", + "accept": false, + "reason": "Different underlying problems in different code paths (`trainer.py`/`trainer_pt_utils.py` vs `generation/logits_process.py`); they are not the same fix and would not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 30, + "component_count": 31, + "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-47315-2", + "nodes": [ + "pull_request:47315", + "pull_request:47430" + ], + "soft_pairs": [ + "pull_request:47315|pull_request:47430" + ], + "prepared_review_unit_hash": "fcb53bfc459fcb5e4e639c2000edb593e5d0969ea6335de551f41e0697c3ebb8", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4382, + "estimated_input_tokens": 1096, + "estimated_eval_tokens": 2448 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4382, + "estimated_input_tokens": 1096, + "estimated_eval_tokens": 2448 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two PRs touch the same compressed_tensors area but fix different bugs: one corrects MoE expert shape reconstruction for non-32-divisible bit widths, the other collects zero_point for asymmetric decompression. They should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: PR 47315 fixes shape reconstruction for non-32-divisible packing, while PR 47430 fixes missing zero_point handling for asymmetric MoE decompression.", + "best_issue_reason": null, + "best_pr_reason": "If one PR must represent the cluster, PR 47315 is the better anchor because it addresses a concrete decompression failure mode with a narrowly scoped fix; however, it is not a duplicate of PR 47430.", + "soft_edge_verdicts": [ + { + "left": "pull_request:47315", + "right": "pull_request:47430", + "accept": false, + "reason": "Different underlying bugs and fixes: shape fallback for low-bit packing vs. zero_point collection for asymmetric quantization. They share a file and subsystem, but not the same code-path problem." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 31, + "component_count": 31, + "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-47400-2", + "nodes": [ + "pull_request:47400", + "pull_request:47421" + ], + "soft_pairs": [ + "pull_request:47400|pull_request:47421" + ], + "prepared_review_unit_hash": "b8517d6f5b6e0ff5cfa2df92dffed83d4325863b9e63b2cc5d43d5ab89afef8c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4618, + "estimated_input_tokens": 1155, + "estimated_eval_tokens": 2566 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4618, + "estimated_input_tokens": 1155, + "estimated_eval_tokens": 2566 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate GGUF model-support PRs touching the same integration code, but they add support for different architectures (Phi-2 vs DeepSeek-V3), so they are not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:47400", + "right": "pull_request:47421", + "accept": false, + "reason": "Same GGUF subsystem and overlapping files, but the concrete changes target different model architectures and different support paths, so they are not the same underlying change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/manifest.json b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..3ef8802a9147172dbd56f0e917b92c3d2c91f607 --- /dev/null +++ b/snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260722t000038z", + "artifacts": { + "hybrid": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-07-22T00:15:46Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260722T000038Z", + "variant": "hybrid" +} diff --git a/snapshots/20260722T000038Z/manifest.json b/snapshots/20260722T000038Z/manifest.json index 60fb0bf9c6ef158d2801f19c3edc3d6fae4b4f62..f9e55a63e0555b690dd0bae87939518f6248ab60 100644 --- a/snapshots/20260722T000038Z/manifest.json +++ b/snapshots/20260722T000038Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 1358 }, "extracted_at": "2026-07-22T00:00:38Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260722t000038z", + "runs": { + "hybrid-model-20260722t000038z": { + "analysis_id": "hybrid-model-20260722t000038z", + "artifacts": { + "hybrid": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260722T000038Z/analysis-runs/hybrid-model-20260722t000038z/manifest.json", + "model": null, + "published_at": "2026-07-22T00:15:46Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260722T000038Z", "watermark": {