diff --git a/analysis/current/analysis-report-hybrid.json b/analysis/current/analysis-report-hybrid.json index 313395d0468703e7a8cc05a531f351402c6219c5..f01dd714caf519ba0073d6aca35173a05cce2f90 100644 --- a/analysis/current/analysis-report-hybrid.json +++ b/analysis/current/analysis-report-hybrid.json @@ -1,8 +1,8 @@ { "schema_version": "1.0", "repo": "huggingface/transformers", - "snapshot_id": "20260608T120020Z", - "generated_at": "2026-06-08T12:17:13Z", + "snapshot_id": "20260608T180018Z", + "generated_at": "2026-06-08T18:11:33Z", "evidence_quality": "full", "llm_enrichment": true, "meta_bugs": [ @@ -1148,6 +1148,364 @@ } ] }, + { + "cluster_id": "cluster-45363-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46339, + "issue_numbers": [], + "pr_numbers": [ + 45363, + 45666, + 46339 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45363, + "right_pr_number": 45666, + "code_similarity": 0.392, + "size_similarity": 0.298, + "file_overlap": 0.571, + "area_overlap": 0.329, + "patch_similarity": 0.306, + "shared_filenames": [ + "fused_qwen_example.py", + "src/transformers/integrations/__init__.py", + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py" + ], + "shared_file_areas": [ + { + "filename": "fused_qwen_example.py", + "left_ranges": [ + [ + 1, + 30 + ] + ], + "right_ranges": [ + [ + 1, + 49 + ] + ] + }, + { + "filename": "src/transformers/integrations/__init__.py", + "left_ranges": [ + [ + 67, + 81 + ], + [ + 226, + 240 + ] + ], + "right_ranges": [ + [ + 67, + 73 + ], + [ + 222, + 228 + ] + ] + }, + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 17 + ], + [ + 22, + 35 + ], + [ + 505, + 716 + ] + ], + "right_ranges": [ + [ + 11, + 53 + ], + [ + 286, + 295 + ], + [ + 521, + 1077 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 3738, + 3747 + ] + ], + "right_ranges": [ + [ + 1330, + 1338 + ], + [ + 3763, + 3789 + ], + [ + 4223, + 4234 + ], + [ + 4628, + 4641 + ] + ] + } + ] + }, + { + "left_pr_number": 45363, + "right_pr_number": 46339, + "code_similarity": 0.506, + "size_similarity": 0.925, + "file_overlap": 0.286, + "area_overlap": 0.545, + "patch_similarity": 0.297, + "shared_filenames": [ + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 17 + ], + [ + 22, + 35 + ], + [ + 505, + 716 + ] + ], + "right_ranges": [ + [ + 14, + 52 + ], + [ + 285, + 294 + ], + [ + 521, + 685 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 3738, + 3747 + ] + ], + "right_ranges": [ + [ + 1346, + 1354 + ], + [ + 3785, + 3811 + ], + [ + 4245, + 4256 + ], + [ + 4650, + 4663 + ] + ] + } + ] + }, + { + "left_pr_number": 45666, + "right_pr_number": 46339, + "code_similarity": 0.399, + "size_similarity": 0.322, + "file_overlap": 0.667, + "area_overlap": 0.247, + "patch_similarity": 0.318, + "shared_filenames": [ + "src/transformers/core_model_loading.py", + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py", + "tests/kernels/test_kernels.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/core_model_loading.py", + "left_ranges": [ + [ + 19, + 25 + ], + [ + 80, + 86 + ] + ], + "right_ranges": [ + [ + 19, + 25 + ], + [ + 80, + 86 + ] + ] + }, + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 53 + ], + [ + 286, + 295 + ], + [ + 521, + 1077 + ] + ], + "right_ranges": [ + [ + 14, + 52 + ], + [ + 285, + 294 + ], + [ + 521, + 685 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 1330, + 1338 + ], + [ + 3763, + 3789 + ], + [ + 4223, + 4234 + ], + [ + 4628, + 4641 + ] + ], + "right_ranges": [ + [ + 1346, + 1354 + ], + [ + 3785, + 3811 + ], + [ + 4245, + 4256 + ], + [ + 4650, + 4663 + ] + ] + }, + { + "filename": "tests/kernels/test_kernels.py", + "left_ranges": [ + [ + 19, + 54 + ], + [ + 228, + 281 + ], + [ + 294, + 317 + ], + [ + 605, + 995 + ] + ], + "right_ranges": [ + [ + 21, + 28 + ], + [ + 33, + 39 + ], + [ + 90, + 103 + ], + [ + 217, + 323 + ], + [ + 336, + 359 + ] + ] + } + ] + } + ] + }, { "cluster_id": "cluster-43240-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #43240.", @@ -1444,11 +1802,11 @@ { "left_pr_number": 46173, "right_pr_number": 46474, - "code_similarity": 0.638, - "size_similarity": 0.424, - "file_overlap": 0.5, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, "area_overlap": 1.0, - "patch_similarity": 0.353, + "patch_similarity": 1.0, "shared_filenames": [ "src/transformers/cache_utils.py" ], @@ -1555,65 +1913,6 @@ } ] }, - { - "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-46291-3", "summary": "Cluster of 1 issues and 2 PRs centered on issue #46291.", @@ -1680,103 +1979,57 @@ ] }, { - "cluster_id": "cluster-46463-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #46463.", + "cluster_id": "cluster-45081-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45081.", "status": "open", - "confidence": 0.55, - "canonical_issue_number": 46463, - "canonical_pr_number": 46465, + "confidence": 0.8, + "canonical_issue_number": 45081, + "canonical_pr_number": 45317, "issue_numbers": [ - 46463 + 45081 ], "pr_numbers": [ - 46465, - 46470 + 45086, + 45317 ], "evidence_types": [ - "closing_reference" + "closing_reference", + "shared_issue_target", + "soft_similarity" ], "pr_comparisons": [ { - "left_pr_number": 46465, - "right_pr_number": 46470, - "code_similarity": 0.501, - "size_similarity": 0.737, - "file_overlap": 0.333, - "area_overlap": 0.647, - "patch_similarity": 0.184, + "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/models/mt5/modeling_mt5.py", - "src/transformers/models/t5/modeling_t5.py", - "src/transformers/models/umt5/modeling_umt5.py" + "src/transformers/tokenization_utils_tokenizers.py" ], "shared_file_areas": [ { - "filename": "src/transformers/models/mt5/modeling_mt5.py", - "left_ranges": [ - [ - 656, - 662 - ], - [ - 665, - 671 - ] - ], - "right_ranges": [ - [ - 654, - 660 - ], - [ - 664, - 670 - ] - ] - }, - { - "filename": "src/transformers/models/t5/modeling_t5.py", - "left_ranges": [ - [ - 645, - 651 - ], - [ - 654, - 660 - ] - ], - "right_ranges": [ - [ - 643, - 649 - ], - [ - 653, - 659 - ] - ] - }, - { - "filename": "src/transformers/models/umt5/modeling_umt5.py", + "filename": "src/transformers/tokenization_utils_tokenizers.py", "left_ranges": [ [ - 614, - 620 + 1360, + 1370 ], [ - 623, - 629 + 1374, + 1380 ] ], "right_ranges": [ [ - 612, - 618 + 1360, + 1370 ], [ - 622, - 628 + 1374, + 1380 ] ] } @@ -1842,131 +2095,6 @@ } ] }, - { - "cluster_id": "cluster-46234-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #46234.", - "status": "open", - "confidence": 0.75, - "canonical_issue_number": 46234, - "canonical_pr_number": 46343, - "issue_numbers": [ - 46234 - ], - "pr_numbers": [ - 46343, - 46475 - ], - "evidence_types": [ - "closing_reference", - "shared_issue_target" - ], - "pr_comparisons": [ - { - "left_pr_number": 46343, - "right_pr_number": 46475, - "code_similarity": 0.619, - "size_similarity": 0.568, - "file_overlap": 0.667, - "area_overlap": 0.727, - "patch_similarity": 0.337, - "shared_filenames": [ - "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", - "src/transformers/models/falcon_mamba/modular_falcon_mamba.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", - "left_ranges": [ - [ - 182, - 190 - ], - [ - 193, - 202 - ] - ], - "right_ranges": [ - [ - 181, - 199 - ] - ] - }, - { - "filename": "src/transformers/models/falcon_mamba/modular_falcon_mamba.py", - "left_ranges": [ - [ - 135, - 143 - ], - [ - 146, - 155 - ] - ], - "right_ranges": [ - [ - 134, - 152 - ] - ] - } - ] - } - ] - }, - { - "cluster_id": "cluster-46424-3", - "summary": "Cluster of 1 issues and 2 PRs centered on issue #46424.", - "status": "open", - "confidence": 0.8, - "canonical_issue_number": 46424, - "canonical_pr_number": 46436, - "issue_numbers": [ - 46424 - ], - "pr_numbers": [ - 46436, - 46481 - ], - "evidence_types": [ - "closing_reference", - "shared_issue_target", - "soft_similarity" - ], - "pr_comparisons": [ - { - "left_pr_number": 46436, - "right_pr_number": 46481, - "code_similarity": 0.59, - "size_similarity": 0.98, - "file_overlap": 0.333, - "area_overlap": 0.615, - "patch_similarity": 0.525, - "shared_filenames": [ - "src/transformers/generation/utils.py" - ], - "shared_file_areas": [ - { - "filename": "src/transformers/generation/utils.py", - "left_ranges": [ - [ - 1918, - 1930 - ] - ], - "right_ranges": [ - [ - 1918, - 1925 - ] - ] - } - ] - } - ] - }, { "cluster_id": "cluster-44018-2", "summary": "Cluster of 2 related pull requests linked by soft_similarity.", @@ -2316,6 +2444,16 @@ "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-45363-3", + "canonical_pr_number": 46339, + "duplicate_pr_numbers": [ + 45363, + 45666 + ], + "target_issue_number": null, + "reason": "PRs in cluster-45363-3 are treated as duplicates because they share soft_similarity evidence." + }, { "cluster_id": "cluster-45561-3", "canonical_pr_number": 45645, @@ -2343,15 +2481,6 @@ "target_issue_number": 45987, "reason": "PRs in cluster-45987-3 are treated as duplicates because they converge on issue #45987 with closing_reference, shared_issue_target evidence." }, - { - "cluster_id": "cluster-46234-3", - "canonical_pr_number": 46343, - "duplicate_pr_numbers": [ - 46475 - ], - "target_issue_number": 46234, - "reason": "PRs in cluster-46234-3 are treated as duplicates because they converge on issue #46234 with closing_reference, shared_issue_target evidence." - }, { "cluster_id": "cluster-46291-3", "canonical_pr_number": 46293, @@ -2360,24 +2489,6 @@ ], "target_issue_number": 46291, "reason": "PRs in cluster-46291-3 are treated as duplicates because they converge on issue #46291 with closing_reference evidence." - }, - { - "cluster_id": "cluster-46424-3", - "canonical_pr_number": 46436, - "duplicate_pr_numbers": [ - 46481 - ], - "target_issue_number": 46424, - "reason": "PRs in cluster-46424-3 are treated as duplicates because they converge on issue #46424 with closing_reference, shared_issue_target, soft_similarity evidence." - }, - { - "cluster_id": "cluster-46463-3", - "canonical_pr_number": 46465, - "duplicate_pr_numbers": [ - 46470 - ], - "target_issue_number": 46463, - "reason": "PRs in cluster-46463-3 are treated as duplicates because they converge on issue #46463 with closing_reference evidence." } ], "best_issue": { diff --git a/analysis/current/analysis-report-hybrid.llm-reviews.json b/analysis/current/analysis-report-hybrid.llm-reviews.json index 7a6194d2c44df1e0e2c35500f2f2a1effe926abf..7a6d68ada704220d50e07725073c84b4cfad2445 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": "20260608T120020Z", - "generated_at": "2026-06-08T12:17:13Z", + "snapshot_id": "20260608T180018Z", + "generated_at": "2026-06-08T18:11:33Z", "model": "gpt-5.4-mini?service_tier=flex", "reviews": [ { @@ -134,430 +134,400 @@ "issue:41762|issue:43596", "issue:39290|issue:41720" ], - "prepared_review_unit_hash": "53e133b87dd14f24cc056edef4409a937eec3291386b5b3cf9cb2f19ff8c201e", + "prepared_review_unit_hash": "ee8bb41fda9a60efc1f33b4f5d6dab030c88af11c8c89b9b9f6e0b4eda4ca225", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 65, - "serialized_chars": 47167, - "estimated_input_tokens": 11792, - "estimated_eval_tokens": 23840 + "serialized_chars": 47171, + "estimated_input_tokens": 11793, + "estimated_eval_tokens": 23842 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous: many unrelated model/runtime/docs regressions were grouped by loose text similarity. The only clear duplicate subclusters are the DeepSpeed ZeRO-3 index-out-of-bounds reports around model init and the Qwen3.5MoE rotary-embedding/offload reports.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43596 is the cleanest representative of the strongest true duplicate subcluster: the repeated ZeRO-3 \"index 0 is out of bounds\" init failure, also seen in 41762 and 43638.", + "summary": "Most pairs are clearly unrelated despite superficial similarity. The only strong duplicate pair is the ZeRO-3/BERT IndexError, and there is one plausible Qwen3_5Moe rotary-embedding device/materialization duplicate pair.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43596 is the cleanest representative of the confirmed ZeRO-3/BERT IndexError duplicate; issue:43638 is the same failure with a more specific phrasing.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43596 is the best issue to anchor triage because it states the shared failure mode succinctly and matches the closest near-duplicate reports.", + "best_issue_reason": "issue:43596 is the best issue to keep: it is well-scoped, generic enough to cover the duplicate, and matches the clearest duplicate partner.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44322", "right": "issue:45464", "accept": false, - "reason": "Different models and failures: Qwen3.5 config attribute error vs chat/completions streaming issue." + "reason": "Different models and failure modes; not the same bug." }, { "left": "issue:45237", "right": "issue:45685", "accept": false, - "reason": "Unrelated problems: AMD GPU support vs MPS histogram on Int." + "reason": "Different backend/model issue; only broad MoE/hardware similarity." }, { "left": "issue:43716", "right": "issue:45237", "accept": false, - "reason": "Different code paths: image preprocessor/model dtype mismatch vs AMD GPU execution failure." + "reason": "Different model-specific runtime errors." }, { "left": "issue:43828", "right": "issue:45237", "accept": false, - "reason": "Both are dtype/device-related at a high level, but they are distinct models and distinct failure modes." + "reason": "Both are dtype/backend problems, but not the same code path." }, { "left": "issue:45237", "right": "issue:46468", "accept": false, - "reason": "Unrelated model-specific example breakage vs AMD GPU support issue." + "reason": "Different model/example failures." }, { "left": "issue:45237", "right": "issue:45507", "accept": false, - "reason": "Different MoE/model method bug vs GPU backend compatibility problem." + "reason": "Different model families and symptoms." }, { "left": "issue:30277", "right": "issue:35545", "accept": false, - "reason": "Different subsystems: DeepSpeed collective mismatch vs ONNX export error." + "reason": "DeepSpeed collective mismatch vs ONNX export error are unrelated." }, { "left": "issue:35545", "right": "issue:42915", "accept": false, - "reason": "No shared concrete bug: ModernBERT ONNX export vs Qwen3Moe FP8 failure." + "reason": "Different model/export issues." }, { "left": "issue:35545", "right": "issue:41720", "accept": false, - "reason": "Different failure modes and model families; not mergeable as one fix." + "reason": "ONNX export vs device-map cuda assert; unrelated." }, { "left": "issue:45901", "right": "issue:46015", "accept": false, - "reason": "One is a task crash, the other is outdated docs/examples; same area but not the same bug." + "reason": "Task crash vs documentation issue; not a duplicate." }, { "left": "issue:44509", "right": "issue:44512", "accept": false, - "reason": "Both are docs cleanup for v5 removals, but they target different removed content and should not be treated as one bug." + "reason": "Both are docs cleanup, but they target different removed references." }, { "left": "issue:35545", "right": "issue:36010", "accept": false, - "reason": "ONNX export bug vs import path breakage; unrelated." + "reason": "Different import/export failures." }, { "left": "issue:44805", "right": "issue:45823", "accept": false, - "reason": "Different tensor-shape error vs device-map forward mismatch." + "reason": "Generic mask-shape error vs Gemma4 device mismatch are not the same bug." }, { "left": "issue:44918", "right": "issue:46165", "accept": false, - "reason": "Different training/trainer issue vs resume/trust-remote-code issue." + "reason": "Different training/resume problems." }, { "left": "issue:43638", "right": "issue:44805", "accept": false, - "reason": "Same broad IndexError wording, but different trigger and subsystem." + "reason": "Different root causes and model paths." }, { - "left": "issue:43596", + "left": "issue:43638", "right": "issue:46165", "accept": false, - "reason": "DeepSpeed ZeRO-3 init bug vs resume/trust-remote-code training issue." + "reason": "ZeRO-3/BERT init error vs trust-remote-code resume issue." }, { "left": "issue:43596", "right": "issue:44805", "accept": false, - "reason": "Generic shape/index errors, but not the same concrete code path." + "reason": "Different IndexError shapes and code paths." }, { "left": "issue:44805", "right": "issue:46294", "accept": false, - "reason": "Different mismatch types and different model/input modalities." + "reason": "Different multimodal input validation failures." }, { "left": "issue:44805", "right": "issue:45161", "accept": false, - "reason": "Mask-shape error vs GPT-OSS MoE TP issue." + "reason": "Mask-shape error vs TP/MoE failure." }, { "left": "issue:46165", "right": "issue:46468", "accept": false, - "reason": "Resume/trust-remote-code training issue is unrelated to the Nemotron example bug." + "reason": "Different model/training issues." }, { "left": "issue:44560", "right": "issue:46165", "accept": false, - "reason": "Video chat template bug vs training resume problem." + "reason": "Video embedding error vs training resume issue." }, { "left": "issue:43596", "right": "issue:46165", "accept": false, - "reason": "Different model-loading/training scenarios and different root causes." + "reason": "Different problems." }, { "left": "issue:34634", "right": "issue:35545", "accept": false, - "reason": "BarkProcessor voice preset bug is unrelated to ModernBERT ONNX export." + "reason": "Processor preset issue vs ONNX export error." }, { "left": "issue:30277", "right": "issue:43638", "accept": false, - "reason": "Both mention DeepSpeed, but one is collective mismatch and the other is ZeRO-3 init indexing; not the same bug." + "reason": "Different ZeRO-3 failures on different models/code paths." }, { "left": "issue:43596", "right": "issue:43638", "accept": true, - "reason": "Same ZeRO-3 index-out-of-bounds init failure on closely related BertModel reports." + "reason": "Same ZeRO-3/BERT IndexError at load/init; only phrasing differs." }, { "left": "issue:43039", "right": "issue:45161", "accept": false, - "reason": "Liger Kernel cross-entropy path vs GPT-OSS tensor-parallel bug." + "reason": "Different features and failure modes." }, { "left": "issue:43975", "right": "issue:44457", "accept": false, - "reason": "Incorrect detokenization vs LoRA merge/save-load mismatch." + "reason": "Detokenization bug vs LoRA save/load mismatch." }, { "left": "issue:44292", "right": "issue:44457", "accept": false, - "reason": "Different symptoms and likely different fix areas." + "reason": "Different model/runtime issue vs LoRA merge mismatch." }, { "left": "issue:24643", "right": "issue:30277", "accept": false, - "reason": "Both are DeepSpeed-related, but one is a weight shape error and the other is a collective mismatch." + "reason": "Different DeepSpeed errors." }, { "left": "issue:34928", "right": "issue:44805", "accept": false, - "reason": "Activation checkpointing/FSDP issue vs mask/tensor shape mismatch." + "reason": "Activation-checkpoint/FSDP recompute mismatch vs mask-shape IndexError." }, { "left": "issue:45859", "right": "issue:45902", "accept": true, - "reason": "Same Qwen3.5MoE rotary-embedding code path; both describe broken buffer/materialization/offload handling." - }, - { - "left": "issue:43638", - "right": "issue:45161", - "accept": false, - "reason": "Different model families and unrelated errors." - }, - { - "left": "issue:34928", - "right": "issue:36331", - "accept": false, - "reason": "Activation checkpointing/FSDP bug vs compute_loss signature breakage." - }, - { - "left": "issue:43596", - "right": "issue:45161", - "accept": false, - "reason": "ZeRO-3 model init bug vs GPT-OSS tensor-parallel bug." - }, - { - "left": "issue:43716", - "right": "issue:45507", - "accept": false, - "reason": "Different model-specific issues: Mistral image dtype mismatch vs GraniteMoE invalid method." - }, - { - "left": "issue:45874", - "right": "issue:46294", - "accept": false, - "reason": "Gemma4 reverse-embedding lookup issue vs Qwen3.5 video template/token mismatch." + "reason": "Same Qwen3_5MoeTextRotaryEmbedding device/materialization bug; one is the CPU-offload symptom, the other the uninitialized inv_freq symptom." }, { "left": "issue:44292", "right": "issue:45237", "accept": false, - "reason": "Different model/runtime failures with no shared code-path." + "reason": "Different model/runtime failures." }, { "left": "issue:39290", "right": "issue:45081", "accept": false, - "reason": "Gemma3/vLLM config attribute issue vs tokenizer regex patch crash." + "reason": "Different model/version regressions." }, { "left": "issue:45823", "right": "issue:45874", "accept": false, - "reason": "Device-map forward mismatch vs expensive reverse embedding lookup; not the same bug." + "reason": "Different Gemma4 issues." }, { "left": "issue:43425", "right": "issue:44292", "accept": false, - "reason": "Torch version incompatibility vs Qwen NVFP4 runtime error." + "reason": "Torch compatibility issue vs model runtime error." }, { "left": "issue:43828", - "right": "issue:45161", + "right": "issue:45685", "accept": false, - "reason": "Autocast dtype mismatch vs GPT-OSS TP issue." + "reason": "Dtype mismatch vs MPS histogram int error are unrelated." }, { "left": "issue:34928", "right": "issue:41720", "accept": false, - "reason": "FSDP checkpointing bug vs Qwen3 auto device mapping assert." + "reason": "Different distributed-training and device-map failures." }, { "left": "issue:43828", "right": "issue:44292", "accept": false, - "reason": "Different dtype/runtime errors on different models." + "reason": "Different model-specific failures." }, { "left": "issue:43541", "right": "issue:43716", "accept": false, - "reason": "Mixtral torch dynamo tracing error vs Mistral image preprocessor dtype mismatch." + "reason": "Different model/tracing vs image-preprocessor dtype issues." }, { "left": "issue:43716", "right": "issue:46468", "accept": false, - "reason": "Different model/example failures." + "reason": "Different model-specific bugs." }, { "left": "issue:43716", "right": "issue:45685", "accept": false, - "reason": "Different backend/device issues and different models." + "reason": "Different device/dtype backends." }, { "left": "issue:44918", "right": "issue:45161", "accept": false, - "reason": "TRL SFT trainer embedding unpacking vs GPT-OSS tensor-parallel failure." + "reason": "Different training feature failures." }, { "left": "issue:44292", "right": "issue:44918", "accept": false, - "reason": "Different failure surfaces and model families." + "reason": "Different model/runtime bugs." }, { "left": "issue:43716", "right": "issue:44292", "accept": false, - "reason": "Different models and mismatch types." + "reason": "Different issues with only loose multimodal overlap." }, { "left": "issue:45507", "right": "issue:45685", "accept": false, - "reason": "GraniteMoE invalid method vs MPS histogram kernel limitation." + "reason": "Different methods/backends and symptoms." }, { "left": "issue:43039", "right": "issue:44918", "accept": false, - "reason": "Liger Kernel cross-entropy call path vs TRL trainer embedding issue." + "reason": "Liger kernel cross-entropy issue vs SFT unpacking error." }, { "left": "issue:41762", "right": "issue:43638", - "accept": true, - "reason": "Same DeepSpeed ZeRO-3 index-out-of-bounds failure pattern, just on different Bert/Gemma3 reports." + "accept": false, + "reason": "Both involve ZeRO-3 IndexError, but on different models and likely different code paths." }, { "left": "issue:43827", "right": "issue:44512", "accept": false, - "reason": "Both are docs-related, but one is about pipeline-task removals and the other about the removed run command." + "reason": "Both are docs regressions, but they cover different removed references." }, { "left": "issue:43828", "right": "issue:45685", "accept": false, - "reason": "Autocast dtype mismatch is unrelated to the MPS Int histogram limitation." + "reason": "Dtype mismatch vs MPS int histogram error are unrelated." }, { "left": "issue:44918", "right": "issue:45685", "accept": false, - "reason": "Training trainer unpacking issue vs MPS backend limitation." + "reason": "Different training/model failures." }, { "left": "issue:43039", "right": "issue:43828", "accept": false, - "reason": "Different mechanisms: Liger Kernel cross_entropy routing vs autocast dtype mismatch." + "reason": "Different bugs." }, { "left": "issue:30277", "right": "issue:36331", "accept": false, - "reason": "Collective mismatch in DeepSpeed vs custom trainer signature breakage." + "reason": "DeepSpeed collective mismatch vs trainer API signature break." }, { "left": "issue:34928", "right": "issue:36010", "accept": false, - "reason": "FSDP/checkpointing issue vs GenerationMixin import error." + "reason": "Different subsystems and failure modes." }, { "left": "issue:45237", "right": "issue:46015", "accept": false, - "reason": "GPU backend/model runtime bug vs documentation example issue." + "reason": "Model/runtime failure vs documentation issue." }, { "left": "issue:43039", "right": "issue:44805", "accept": false, - "reason": "Liger Kernel cross-entropy routing vs tensor-mask shape mismatch." + "reason": "Different code paths and symptoms." }, { "left": "issue:43866", "right": "issue:46402", "accept": false, - "reason": "Corrupted checkpoint vs model weights not loaded; different problems." + "reason": "Checkpoint corruption vs missing weight loading." }, { "left": "issue:45800", "right": "issue:46304", "accept": false, - "reason": "Torch/Transformers version incompatibility vs distributed build support regression." + "reason": "Different compatibility regressions." }, { "left": "issue:43296", "right": "issue:46304", "accept": false, - "reason": "PaddleOCR-VL/vLLM loading issue vs USE_DISTRIBUTED=0 regression." + "reason": "Different release incompatibilities." }, { "left": "issue:41762", "right": "issue:43596", - "accept": true, - "reason": "Same ZeRO-3 index-out-of-bounds init failure class." + "accept": false, + "reason": "Same broad ZeRO-3 error shape, but different models and contexts." }, { "left": "issue:39290", "right": "issue:41720", "accept": false, - "reason": "Different model/version regressions: Gemma3/vLLM config attr vs Qwen3 auto device-map assert." + "reason": "Different model/version-specific failures." } ] }, "evaluator_result": { "accept": true, - "feedback": "The triage summary is grounded in the packet, and the soft-edge verdicts stay conservative overall. The only accepted pairs are well-supported by the titles and shared failure class (the ZeRO-3 index-out-of-bounds reports and the Qwen3.5MoE rotary-embedding/offload pair), while the rejections appropriately avoid over-claiming duplicate status for loosely similar but distinct issues." + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The two accepted duplicates are consistent with the titles/body signals: 43596\u219443638 and 45859\u219445902. The rejections are also appropriately cautious and do not overclaim duplication from superficial similarity." }, "evaluator_used": true, "retried": false, @@ -671,283 +641,253 @@ "issue:43531|issue:45070", "issue:45020|issue:45800" ], - "prepared_review_unit_hash": "39a6a802d4745e5215230cb4c7915dee9b13e38260c6e3f97067090d475de01a", + "prepared_review_unit_hash": "d1c1b356f5ad4477df0f519a94c4e725c2085e8565289da3df458ca723d25a6e", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 42, - "serialized_chars": 40093, - "estimated_input_tokens": 10024, - "estimated_eval_tokens": 20304 + "serialized_chars": 40100, + "estimated_input_tokens": 10025, + "estimated_eval_tokens": 20306 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issues; the only strong duplicate-looking pair is the v5 AutoTokenizer wrong-token-id regression across model families. No PRs were provided.", - "confidence": 0.87, - "canonical_issue_reason": "Issue 45920 is the broadest, most representative tokenizer regression: it describes wrong token IDs across multiple model families and explicitly frames it as a v5 regression.", + "summary": "The cluster is dominated by one clear umbrella issue: the v5 AutoTokenizer wrong-token-ID regression. Most other pairs are unrelated single-model, docs, or compatibility bugs. The DeepSpeed ZeRO-3/Bert index-error pair is now treated conservatively as non-duplicate because the shared symptom is too generic to prove one root cause.", + "confidence": 0.83, + "canonical_issue_reason": "issue:45920 is the broadest and most representative tokenizer regression report: it spans multiple affected model families, is still open, and has inbound references, making it the best umbrella issue for the duplicate cluster.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45920 is the best canonical candidate overall because it has the widest scope, clear regression framing, and is the most likely umbrella for the tokenizer duplicate cluster.", + "best_issue_reason": "issue:45920 is the best cluster representative because it captures the widest scope of the tokenizer-ID regression across several models, not just one family.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44589", "right": "issue:46304", "accept": false, - "reason": "Different failures: float8 storage lookup vs distributed-mode support; no shared code path." + "reason": "Different failures: a missing storage class vs a torch/distributed compatibility regression." }, { "left": "issue:45823", "right": "issue:46294", "accept": false, - "reason": "Both are model-specific runtime issues, but they affect different components and code paths (device_map forward mismatch vs chat template video feature mismatch)." + "reason": "Different model paths and symptoms: Gemma4 device mismatch vs Qwen3.5 video/chat-template token mismatch." }, { "left": "issue:43901", "right": "issue:44512", "accept": false, - "reason": "Both are docs-related, but they cover different documentation errors and different removed/changed behaviors." + "reason": "Both are docs issues, but they concern unrelated documentation mismatches." }, { "left": "issue:45237", "right": "issue:45901", "accept": false, - "reason": "Unrelated subsystems and symptoms: AMD GPU support for GPT-OSS vs table-question-answering task crash." + "reason": "Different subsystems and failure modes: AMD GPU support vs table QA task crash." }, { "left": "issue:44841", "right": "issue:45823", "accept": false, - "reason": "Different models and different runtime failures; no evidence they stem from the same bug." + "reason": "Different models and bugs: Voxtral processor failure vs Gemma4 device placement issue." }, { "left": "issue:43039", "right": "issue:44560", "accept": false, - "reason": "Liger Kernel cross-entropy dispatch and Qwen3.5-VL video StopIteration are unrelated." + "reason": "Liger kernel cross-entropy routing and Qwen3-VL video StopIteration are unrelated." }, { "left": "issue:43975", "right": "issue:46015", "accept": false, - "reason": "Different problem types: detokenization regression vs outdated RAG examples." + "reason": "Model detokenization bug vs outdated RAG docs; not the same underlying issue." }, { "left": "issue:44863", "right": "issue:46015", "accept": false, - "reason": "Model loading/architecture issue vs documentation drift; not the same underlying bug." + "reason": "NemotronH checkpoint-loading failure is unrelated to stale RAG examples." }, { "left": "issue:45020", "right": "issue:46304", "accept": false, - "reason": "Both mention recent-version regressions, but one is about remote_code models and the other about USE_DISTRIBUTED=0 compatibility." + "reason": "Remote_code breakage and USE_DISTRIBUTED=0 support are different compatibility problems." }, { "left": "issue:45812", "right": "issue:45920", "accept": true, - "reason": "Both report v5 tokenizer regressions producing wrong token IDs across multiple model families; they look like the same underlying bug and could be fixed together." + "reason": "Same underlying tokenizer regression: both report wrong token IDs from AutoTokenizer in v5 across multiple models." }, { "left": "issue:43866", "right": "issue:44863", "accept": false, - "reason": "Different model-loading failures with different root causes; one says corrupted checkpoint, the other says the implementation cannot load checkpoints." + "reason": "Ovis2 checkpoint corruption and NemotronH loading failure are unrelated." }, { "left": "issue:44292", "right": "issue:46015", "accept": false, - "reason": "NVFP4 model-running error is unrelated to outdated RAG examples." + "reason": "A model loading error and a docs issue do not share the same bug." }, { "left": "issue:43872", "right": "issue:46304", "accept": false, - "reason": "bitsandbytes constructor incompatibility is unrelated to distributed build support." + "reason": "bitsandbytes constructor incompatibility is unrelated to distributed support changes." }, { "left": "issue:39290", "right": "issue:43531", "accept": false, - "reason": "Both involve sliding-window wording, but they concern different models and different failures; not enough evidence of one bug." + "reason": "Both mention sliding_window, but they affect different model families and look like separate bugs." }, { "left": "issue:41093", "right": "issue:43638", "accept": false, - "reason": "Both are IndexErrors, but the exact failing condition and model/setup differ significantly." + "reason": "Both are DeepSpeed ZeRO-3/Bert index errors, but the shared out-of-bounds symptom is too generic to confirm one concrete root cause." }, { "left": "issue:43792", "right": "issue:45901", "accept": false, - "reason": "Whisper inference failure and table QA crash are unrelated." + "reason": "Whisper runtime failure and table QA crash are unrelated." }, { "left": "issue:43854", "right": "issue:43866", "accept": false, - "reason": "Both are model-loading complaints, but they concern different models and different failure modes." + "reason": "Different model-loading problems with no shared underlying bug." }, { "left": "issue:43425", "right": "issue:45070", "accept": false, - "reason": "Different compatibility regressions: Torch 2.10 incompatibility vs a pydantic/PretrainedConfig break." + "reason": "Torch version incompatibility and pydantic config breakage are different regression types." }, { "left": "issue:43596", "right": "issue:45800", "accept": false, - "reason": "Different versioning problems with different symptoms; one is a ZeRO-3 index error, the other a torch/transformers incompatibility." + "reason": "ZeRO-3/Bert index error is unrelated to torch 2.4.1 vs transformers 5.8.0 compatibility." }, { "left": "issue:43975", "right": "issue:45901", "accept": false, - "reason": "Detokenization regression and table QA crash are unrelated." + "reason": "Detokenization bug and table QA crash are unrelated." }, { "left": "issue:43792", "right": "issue:45070", "accept": false, - "reason": "Whisper model runtime failure is unrelated to the pydantic config regression." + "reason": "Whisper loading/runtime failure is unrelated to PretrainedConfig pydantic-model changes." }, { "left": "issue:45362", "right": "issue:45464", "accept": false, - "reason": "Both involve Qwen3.5, but one is a transformers chat crash and the other is a streaming chat/completions API failure; different code paths." + "reason": "Same model family, but different surfaces and insufficient evidence of one concrete shared bug." }, { - "left": "issue:44292", + "left": "issue:45685", "right": "issue:45901", "accept": false, - "reason": "Unrelated model-running issue vs task crash." - }, - { - "left": "issue:43404", - "right": "issue:43866", - "accept": false, - "reason": "Weight-tying bug in Mistral3 is unrelated to Ovis2 checkpoint corruption." - }, - { - "left": "issue:45464", - "right": "issue:45685", - "accept": false, - "reason": "Qwen streaming inference and MPS histogram integer support are unrelated." - }, - { - "left": "issue:43541", - "right": "issue:45685", - "accept": false, - "reason": "Different models, different operators, and different backends; no shared concrete bug." - }, - { - "left": "issue:43541", - "right": "issue:43828", - "accept": false, - "reason": "Both mention MoE and numeric issues, but one is grouped_mm tracing and the other is an autocast dtype mismatch; not the same bug." - }, - { - "left": "issue:44805", - "right": "issue:44918", - "accept": false, - "reason": "Different failures in different training/loading contexts; no strong evidence of a shared root cause." - }, - { - "left": "issue:36010", - "right": "issue:39290", - "accept": false, - "reason": "Import path regression for GenerationMixin vs Gemma3 sliding_window_pattern attribute error; unrelated." + "reason": "MPS MoE histogram error and table QA crash are unrelated." }, { "left": "issue:45800", "right": "issue:45823", "accept": false, - "reason": "General torch/transformers compatibility issue vs a Gemma4 device placement bug." + "reason": "Torch version incompatibility and Gemma4 device mismatch are separate issues." }, { "left": "issue:41093", "right": "issue:41762", "accept": false, - "reason": "Both are Zero-3/IndexError themed, but the exact error and trigger differ." + "reason": "Both are ZeRO-3/Bert-related index errors, but the concrete failures differ enough that they do not clearly collapse to one bug." }, { "left": "issue:43296", "right": "issue:43366", "accept": false, - "reason": "Different model families and different loading problems; one is PaddleOCR-VL in vLLM, the other is GGUF gpt-oss support." + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are different issues." }, { "left": "issue:43572", "right": "issue:45070", "accept": false, - "reason": "Different config-field regressions; no evidence they are the same defect." + "reason": "Config-field absence in StableLm and a Pydantic PretrainedConfig break are related only at a very broad level." }, { "left": "issue:44805", "right": "issue:44918", "accept": false, - "reason": "Mask shape mismatch vs embedding unpacking failure are distinct bugs." + "reason": "They both fail with index-related errors, but one is a generic mask-shape mismatch and the other is a trainer embedding unpacking failure." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Both involve MoE models, but the grouped_mm tracing error and TP-only GPT-OSS issue are not the same bug." }, { "left": "issue:44387", "right": "issue:45005", "accept": false, - "reason": "Both are v5 regressions, but one is int4 memory/OOM and the other is tied-weights translation behavior; not the same fix." + "reason": "Int4 OOM and tied-weights translation issues are unrelated regressions." }, { "left": "issue:43828", "right": "issue:45464", "accept": false, - "reason": "Autocast dtype mismatch is unrelated to Qwen3.5 streaming API failures." + "reason": "Autocast dtype mismatch and Qwen3.5 streaming API failure are different failures." }, { "left": "issue:44291", "right": "issue:45070", "accept": false, - "reason": "Unexpected _is_hf_initialized argument during init_empty_weights is a different breakage than the pydantic PretrainedConfig field issue." + "reason": "init_empty_weights/_is_hf_initialized TypeError is unrelated to the pydantic PretrainedConfig regression." }, { "left": "issue:43366", "right": "issue:45464", "accept": false, - "reason": "GGUF gpt-oss support and Qwen3.5 streaming errors are unrelated." + "reason": "GGUF gpt-oss support and Qwen3.5 streaming chat failure are unrelated." }, { "left": "issue:43531", "right": "issue:45070", "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to a pydantic config regression." + "reason": "Sliding-window model config issue and PretrainedConfig pydantic break are different bugs." }, { "left": "issue:45020", "right": "issue:45800", "accept": false, - "reason": "Both are version-regression reports, but they affect different compatibility boundaries and code paths." + "reason": "Remote_code regression and torch/transformers version incompatibility are separate compatibility problems." } ] }, "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet: issue 45920 is the clearest umbrella tokenizer regression, and the only explicitly duplicate-looking pair in the packet is 45812/45920. The soft-edge verdicts are conservative and avoid overstating weak similarity. No PR pairs were present to validate mergeability." + "accept": false, + "feedback": "The overall representative issue choice is grounded, especially issue:45920 as the broadest tokenizer-ID regression report. However, the soft-edge acceptance of issue:41093 \u2194 issue:43638 is too strong for the packet: both are DeepSpeed/ZeRO-3/Bert-related index errors, but the titles only show broadly similar failure types, not a clearly shared root cause. That makes the verdict insufficiently conservative. The other soft-edge rejections look fine." }, "evaluator_used": true, - "retried": false, + "retried": true, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1047,20 +987,20 @@ "issue:45920|issue:46489", "issue:43950|issue:45902" ], - "prepared_review_unit_hash": "370f029876f051fe5ac79ffa765debaf38aec787aa13eee1cd0a2b0bd22b20e5", + "prepared_review_unit_hash": "deff84affc955a481cfb30bd309559626bc13c7a5418af49e02e128eb1f31b46", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 32, - "serialized_chars": 36860, + "serialized_chars": 36859, "estimated_input_tokens": 9215, "estimated_eval_tokens": 18686 }, @@ -1069,213 +1009,213 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous. The only clear duplicate family is the v5 non-persistent-buffer corruption regression; a few other pairs are superficially similar but don\u2019t look like the same concrete bug from titles alone.", - "confidence": 0.63, - "canonical_issue_reason": "issue:43950 is the cleanest anchor for the non-persistent-buffer regression because it explicitly names from_pretrained and register_buffer(persistent=False) corruption; issue:44534 is a close wording variant, but 43950 is more diagnostic.", + "summary": "Mostly a mixed set of unrelated Transformers bug reports. The only clear duplicate cluster here is the v5 non-persistent-buffer corruption regression, where 43644 and 44534 match 43950 closely.", + "confidence": 0.88, + "canonical_issue_reason": "43950 is the clearest representative of the non-persistent-buffer regression: it states the v5.x corruption bug directly and matches the other clear duplicate reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43950 is the strongest representative of the core duplicate family, since it isolates the exact regression mechanism and symptom rather than a broader paraphrase.", + "best_issue_reason": "43950 is the best single issue to anchor this cluster because it is the most specific and directly describes the shared bug seen in the duplicate pair.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43716", "right": "issue:46294", "accept": false, - "reason": "Different model paths and different failures: image-preprocessor dtype mismatch vs video chat-template token/feature mismatch." + "reason": "Different concrete failures: image/text dtype mismatch vs video chat-template token/feature mismatch." }, { "left": "issue:43541", "right": "issue:44589", "accept": false, - "reason": "Unrelated runtime errors in different code paths; only a broad model-execution context is shared." + "reason": "Unrelated runtime paths: torch dynamo grouped_mm tracing error vs float8 storage lookup failure." }, { "left": "issue:44479", "right": "issue:46294", "accept": false, - "reason": "Both involve Qwen video input, but the concrete bugs differ: broad v5.3.0 regression vs chat_template feature/token mismatch." + "reason": "Both involve Qwen video input, but the reported symptoms and code paths are different enough that this is not a safe duplicate." }, { "left": "issue:41093", "right": "issue:43596", "accept": false, - "reason": "Both are index/shape errors with ZeRO-3-style workflows, but the specific failures and model paths differ." + "reason": "Both are index/shape errors around init, but they affect different models and failure points." }, { "left": "issue:43054", "right": "issue:43994", "accept": false, - "reason": "Both concern SigLIP2 behavior, but one is text-embedding degradation and the other is nonsensical AutoModel/pipeline output; not enough to call them the same bug." + "reason": "Same model family, but one is a text-embedding quality complaint and the other is a broader pipeline/AutoModel output issue." }, { "left": "issue:45759", "right": "issue:45902", "accept": false, - "reason": "Different subsystems: multimodal config unwrapping vs non-persistent buffer materialization/uninitialized memory." + "reason": "Different bugs: composite config unwrapping vs uninitialized rotary embedding memory after materialization." }, { "left": "issue:45542", "right": "issue:45800", "accept": false, - "reason": "Unrelated environment/version issues; no shared code-path bug." + "reason": "Completely unrelated: TensorFlow backend detection vs torch/transformers version incompatibility." }, { "left": "issue:43716", "right": "issue:44918", "accept": false, - "reason": "Different modalities and failures: image dtype mismatch vs TRL SFT embedding unpacking." + "reason": "Different models and different multimodal/trainer failure modes." }, { "left": "issue:39186", "right": "issue:44918", "accept": false, - "reason": "Both mention training/runtime problems, but the reported errors and model stacks are unrelated." + "reason": "FSDP weight-shape error is unrelated to TRL embedding unpacking." }, { "left": "issue:45632", "right": "issue:45698", "accept": false, - "reason": "Both involve custom/loading behavior, but one is a cache-path collision and the other is wrong module loading after save/load; not the same bug." + "reason": "Both concern custom-module loading, but one is a leaf-dir cache collision and the other is wrong module selection after save/load; not the same bug." }, { "left": "issue:43645", "right": "issue:45399", "accept": false, - "reason": "Notebook custom-model initialization and flash-attn fallback gating are separate issues." + "reason": "Jupyter custom model initialization issue is unrelated to flash-attn fallback gating." }, { "left": "issue:43404", "right": "issue:45005", "accept": false, - "reason": "Related to tied weights in a broad sense, but the model families and concrete failures are different." + "reason": "Specific Mistral3 weight-tying bug vs a broader translation-model tied-weights regression; too different to merge." }, { "left": "issue:40990", "right": "issue:44960", "accept": false, - "reason": "Different models and symptoms; both are quality/performance reports, not the same underlying bug." + "reason": "No meaningful overlap beyond being model-quality reports." }, { "left": "issue:42915", "right": "issue:45399", "accept": false, - "reason": "FP8 config failure vs flash-attn fallback gating are unrelated." + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn fallback checks." }, { "left": "issue:44530", "right": "issue:45005", "accept": false, - "reason": "PagedAttentionCache/group-type crash is unrelated to tied-weight behavior." + "reason": "PagedAttentionCache group-type crash is unrelated to tied-weight regressions." }, { "left": "issue:43873", "right": "issue:45005", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to tied weights." + "reason": "Quantization/offloading behavior is unrelated to tied-weight handling." }, { "left": "issue:44360", "right": "issue:44512", "accept": false, - "reason": "A code/algorithm discussion issue and a documentation typo are not duplicates." + "reason": "An algorithmic DSA/indexer complaint is unrelated to a docs typo." }, { "left": "issue:45183", "right": "issue:46463", "accept": false, - "reason": "They live in the same input-validation/error-message area, but titles alone do not establish the same concrete bug; this could be a separate model-family-specific messaging issue." + "reason": "Related validation area, but not the same concrete bug or code path." }, { "left": "issue:46489", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and describe different concrete round-trip/encoding failures." + "reason": "Different tokenizer regressions for different model families." }, { "left": "issue:44863", "right": "issue:45005", "accept": false, - "reason": "Model checkpoint loading failure vs tied-weight issue are unrelated." + "reason": "Checkpoint-loading failure is unrelated to tied-weights regression." }, { "left": "issue:44918", "right": "issue:45507", "accept": false, - "reason": "Different model families and different failure modes." + "reason": "Different model families and unrelated failure modes." }, { "left": "issue:44492", "right": "issue:44512", "accept": false, - "reason": "A typo and a removed-command docs issue are separate documentation problems." + "reason": "Unrelated typo vs v5 docs command removal." }, { "left": "issue:43975", "right": "issue:44292", "accept": false, - "reason": "Different models and different tokenizer/runtime failures." + "reason": "Both are tokenizer/model issues, but they concern different models and different symptoms." }, { "left": "issue:44292", "right": "issue:45507", "accept": false, - "reason": "No shared code-path or symptom." + "reason": "Different model-specific runtime failures." }, { "left": "issue:43950", "right": "issue:44534", "accept": true, - "reason": "These describe the same non-persistent-buffer corruption regression in v5, just with different wording." + "reason": "Same underlying v5 regression where non-persistent buffers get corrupted/fill with junk during load." }, { "left": "issue:39692", "right": "issue:43054", "accept": false, - "reason": "Docs example errors are not the same bug as the observed SigLIP2 embedding quality issue." + "reason": "Both mention SigLIP2, but one is a docs/example error report and the other is a model-quality regression." }, { "left": "issue:43644", "right": "issue:43950", "accept": true, - "reason": "Same underlying v5 regression: non-persistent buffers are corrupted or junk-filled on load/materialization." + "reason": "These report the same non-persistent-buffer corruption regression in Transformers v5, phrased slightly differently." }, { "left": "issue:36331", "right": "issue:41762", "accept": false, - "reason": "Different trainer API mismatch vs ZeRO-3 initialization index error." + "reason": "Different errors: custom trainer kwarg mismatch vs ZeRO-3 load-time index error." }, { "left": "issue:43828", "right": "issue:44928", "accept": false, - "reason": "Both mention dtype/precision issues, but the concrete mechanisms and failures differ substantially." + "reason": "Both are mixed-precision/dtype-related, but the models, symptoms, and code paths differ." }, { "left": "issue:43872", "right": "issue:44589", "accept": false, - "reason": "Bitsandbytes constructor incompatibility vs missing Float8 storage object are unrelated." + "reason": "Both are quantization/storage compatibility issues, but they are distinct TypeErrors with different root causes." }, { "left": "issue:45920", "right": "issue:46489", "accept": false, - "reason": "Broad tokenizer regression across many models vs a DeepSeek-Coder v1-specific detokenization problem; not enough evidence they are the same bug." + "reason": "Broader tokenizer regression vs a specific DeepSeek-Coder v1 tokenization bug; not enough evidence they are the same defect." }, { "left": "issue:43950", "right": "issue:45902", "accept": false, - "reason": "Both involve buffers/state, but one is non-persistent-buffer corruption and the other is uninitialized rotary embedding memory after to_empty materialization." + "reason": "Non-persistent-buffer corruption is unrelated to the Qwen3_5Moe rotary-embedding memory bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster summary is mostly conservative, but I can\u2019t accept the canonical/best-issue rationale as written. Issue 43950 is clearly in the non-persistent-buffer regression family, but the packet does not support calling it the \u201cbroadest\u201d title in that subtopic; issue 44534 is at least as broad or broader on its face. So the selection is plausible, but the justification overstates the evidence. The 45183/46463 pair also reads as a plausible duplicate rather than something firmly established without more body evidence." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate set is the non-persistent-buffer regression around 43644/44534/43950, and the other soft-edge verdicts are mostly cautious false negatives rather than overclaims." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -1374,20 +1314,20 @@ "issue:43742|issue:45399", "issue:43404|issue:43950" ], - "prepared_review_unit_hash": "9019cca142b470fc62facbe5862d43d0114dbd7395a4d93828328526170724d3", + "prepared_review_unit_hash": "d5000620e0e3b0a15de50f4531aa0b97932ea6adaebaf3473fa39e939da89b95", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36738, + "serialized_chars": 36740, "estimated_input_tokens": 9185, "estimated_eval_tokens": 18626 }, @@ -1396,198 +1336,198 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab bag of model-specific loading/tokenization/memory regressions, not true duplicates. The soft-similarity pairs share broad themes but not the same concrete bug path, so I reject them all.", - "confidence": 0.78, - "canonical_issue_reason": "issue:45020 is the broadest umbrella report in the set: it frames a general v5 regression breaking remote_code-based model loading, which is closer to the overall theme than the more model-specific reports.", + "summary": "This cluster is mostly a grab bag of model-loading/tokenization regressions plus a few unrelated performance/configuration bugs; there is no true duplicate group. If one issue must serve as the anchor, 45020 is the broadest umbrella report.", + "confidence": 0.56, + "canonical_issue_reason": "issue:45020 is the most general and central report: it describes a broad regression where recent transformers versions break remote_code model loading, which overlaps conceptually with several model-specific loading failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45020 is the best representative issue because it is the most general user-facing regression and can contextualize several of the narrower loading/tokenizer complaints.", + "best_issue_reason": "issue:45020 is the best representative issue because it is the broadest umbrella symptom and most likely to subsume related version-regression reports.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43866", "right": "issue:44403", "accept": false, - "reason": "Unrelated symptoms: corrupted checkpoint vs loading noise. No shared code path." + "reason": "Checkpoint corruption during model load is unrelated to a logging/noise complaint." }, { "left": "issue:38175", "right": "issue:43054", "accept": false, - "reason": "Both mention SigLIP2, but one is zero probabilities and the other is an embedding-quality discrepancy; not the same bug." + "reason": "Both involve SigLIP/SigLIP2, but one is zero probabilities and the other is embedding quality; different bugs." }, { "left": "issue:41553", "right": "issue:43883", "accept": false, - "reason": "Different model families and failure modes: bad tokenizer error message vs missing all_tied_weights_keys." + "reason": "Both are loading errors, but Voxtral tokenizer error and Molmo missing attribute are distinct code paths." }, { "left": "issue:43054", "right": "issue:45701", "accept": false, - "reason": "One is a SigLIP2 model-quality issue, the other is a general tokenization-version change report." + "reason": "Text-embedding quality regression is not the same as a broad tokenization-version change report." }, { "left": "issue:44910", "right": "issue:45859", "accept": false, - "reason": "Different mechanisms: flash-attn packed-sequence/position_ids crash vs CPU-offload incompatibility in rotary embedding." + "reason": "Flash-Attention crash from 3D position_ids is different from CPU-offload incompatibility in rotary embedding." }, { "left": "issue:45663", "right": "issue:45902", "accept": false, - "reason": "Both are Gemma/Qwen-style training issues, but one is an FSDP2 shared_kv_states bug and the other is a meta->to_empty buffer materialization bug." + "reason": "Both mention Gemma4/Qwen3 MoE internals, but FSDP2 KeyError and uninitialized memory are separate defects." }, { "left": "issue:42947", "right": "issue:44387", "accept": false, - "reason": "PEFT gradient checkpointing inefficiency and int4 reserved-memory regression are separate performance issues." + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to int4 quantization memory growth." }, { "left": "issue:41553", "right": "issue:43881", "accept": false, - "reason": "Different models and unrelated load failures; only the generic 'loading failed' theme matches." + "reason": "Voxtral tokenizer loading and glm-4v-9b loading failure are different model-specific issues." }, { "left": "issue:44403", "right": "issue:44451", "accept": false, - "reason": "Generic loading noise vs inability to load a specific model; no evidence they stem from the same defect." + "reason": "Generic load-time noise is not the same as failing to load a specific model." }, { "left": "issue:24643", "right": "issue:28530", "accept": false, - "reason": "Training DeepSpeed weight-shape RuntimeError and early-stopping metric mismatch are unrelated." + "reason": "DeepSpeed weight-shape runtime error is unrelated to early stopping metric resolution." }, { "left": "issue:43054", "right": "issue:43493", "accept": false, - "reason": "Related to SigLIP2, but one reports worse embeddings and the other an HF-vs-JAX discrepancy; not enough to treat as the same bug." + "reason": "Both concern SigLIP2, but one is embedding quality and the other is implementation discrepancy against JAX." }, { "left": "issue:43975", "right": "issue:45520", "accept": false, - "reason": "Tokenizer detokenization bug and Python 3.13 flash_attn import KeyError are unrelated." + "reason": "Tokenizer detokenization bug and Python 3.13 flash_attn import failure are unrelated." }, { "left": "issue:44403", "right": "issue:44488", "accept": false, - "reason": "Both are loading complaints, but one is generic noise and the other is a specific model load failure." + "reason": "Load-time noise and a specific tokenizer load failure for cjvt/sleng-bert are not the same bug." }, { "left": "issue:42947", "right": "issue:43856", "accept": false, - "reason": "Different MoE-related performance issues: gradient checkpointing vs memory usage under training." + "reason": "LoRA checkpointing behavior and MoE training memory usage are separate performance issues." }, { "left": "issue:43643", "right": "issue:45020", "accept": false, - "reason": "One is missing fields from AutoConfig with trust_remote_code, the other is broad model loading regressions; related theme, not same bug." + "reason": "Missing fields in AutoConfig with trust_remote_code is related to remote-code handling but not the same concrete failure as broad model-loading breakage." }, { "left": "issue:41762", "right": "issue:42915", "accept": false, - "reason": "Both involve loading on special runtimes, but Gemma3 ZeRO-3 index error and Qwen3 MoE FP8 failure are distinct code paths." + "reason": "Gemma3 ZeRO-3 loading crash and Qwen3Moe FineGrainedFP8Config failure are different model/config combinations." }, { "left": "issue:45685", "right": "issue:45823", "accept": false, - "reason": "MPS histogram integer error and Gemma4 device mismatch are different backend/model issues." + "reason": "An MPS histogram op error in MoE is unrelated to Gemma4 device placement mismatch." }, { "left": "issue:43881", "right": "issue:43883", "accept": false, - "reason": "Both are load failures, but one is glm-4v-9b and the other is Molmo attribute error; no shared root cause." + "reason": "glm-4v-9b loading failure and Molmo missing all_tied_weights_keys are distinct loader bugs." }, { "left": "issue:44991", "right": "issue:45399", "accept": false, - "reason": "Tokenizer loading regression and flash-attn fallback gating are unrelated." + "reason": "Tokenizer loading for a specific model and flash-attn fallback selection are unrelated." }, { "left": "issue:43824", "right": "issue:43872", "accept": false, - "reason": "ImportError for a model class and bitsandbytes Int8Params incompatibility are separate problems." + "reason": "Missing Qwen2.5-VL import and bitsandbytes Int8Params incompatibility are different integration failures." }, { "left": "issue:45020", "right": "issue:45823", "accept": false, - "reason": "General remote_code loading regressions do not match a Gemma4 device-map forward mismatch." + "reason": "Remote_code loading regressions are not the same as Gemma4 forward device-map mismatch." }, { "left": "issue:44661", "right": "issue:45020", "accept": false, - "reason": "Tokenizer mapping-name failure and remote_code loading regression are different failure points." + "reason": "A tokenizer-mapping edge case in add-new-model-like is narrower and different from the broader remote_code loading regression." }, { "left": "issue:43828", "right": "issue:45823", "accept": false, - "reason": "Autocast dtype mismatch in Phi-tiny-MoE is unrelated to Gemma4 device placement." + "reason": "Dtype mismatch under autocast and Gemma4 device mismatch are separate runtime placement issues." }, { "left": "issue:42491", "right": "issue:44368", "accept": false, - "reason": "Both reference Qwen3/Qwen3.5 LoRA usage, but one is a version-compatibility bug and the other is a warning about tie_word_embeddings." + "reason": "Qwen3 MoE LoRA compatibility across versions is not the same as a warning about tie_word_embeddings in a different Qwen model." }, { "left": "issue:45081", "right": "issue:45399", "accept": false, - "reason": "Tokenizer regex crash and flash-attn fallback blocking are different subsystems." + "reason": "Mistral regex patch crash and flash-attn fallback gating are unrelated tokenizer vs attention-selection bugs." }, { "left": "issue:45715", "right": "issue:46315", "accept": false, - "reason": "Both are tokenizer performance regressions, but one is convert_ids_to_tokens and the other convert_tokens_to_ids; different hot paths." + "reason": "Both are tokenizer performance issues, but one is convert_ids_to_tokens and the other is convert_tokens_to_ids; not the same path." }, { "left": "issue:45488", "right": "issue:46489", "accept": false, - "reason": "Both involve DeepSeek tokenization, but the reported root causes differ: tokenizer pre-tokenizer override vs a v5 detokenization regression." + "reason": "Both involve DeepSeek tokenizers, but this is a ByteLevel/Metaspace override issue while the other is an output regression from a different fix gap." }, { "left": "issue:43782", "right": "issue:44960", "accept": false, - "reason": "Qwen3VL weight_only loading error and a standalone GLM5 issue are unrelated." + "reason": "Qwen3VL weight_only loading error and GLM5 are unrelated model-load problems." }, { "left": "issue:45020", "right": "issue:45356", "accept": false, - "reason": "Broad remote_code loading regression vs a specific Kimi-K2.5 tokenizer codec/fix_mistral_regex regression; different concrete bug." + "reason": "Remote_code breakage is broader than the specific Kimi-K2.5 tokenizer codec and warning regression." }, { "left": "issue:43742", "right": "issue:45399", "accept": false, - "reason": "MobileLLM key error and flash-attn fallback gating are unrelated failures." + "reason": "MobileLLM key error and flash-attn fallback selection do not share the same underlying bug." }, { "left": "issue:43404", "right": "issue:43950", "accept": false, - "reason": "Missing lm_head tying in Mistral3 and non-persistent buffer corruption are distinct state-loading bugs." + "reason": "A tied lm_head issue in Mistral3 is not the same as corrupting non-persistent buffers during from_pretrained." } ] }, @@ -1691,232 +1631,208 @@ "issue:45812|issue:46489", "issue:44560|issue:44918" ], - "prepared_review_unit_hash": "63eee3e75887a09fbc60393e5c89a67d70ad27081c5c7f3d5bebd95c85dd4a29", + "prepared_review_unit_hash": "7875b63d595d524e8f7606193418dba56e624bccb4357811c9c330c7eeb3d97c", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36648, - "estimated_input_tokens": 9162, - "estimated_eval_tokens": 18580 + "serialized_chars": 36649, + "estimated_input_tokens": 9163, + "estimated_eval_tokens": 18582 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly distinct issues; only a few soft pairs look like true duplicates, centered on Transformers 5.x custom-model init and the `_is_hf_initialized` constructor regression, plus the v5 docs pipeline cleanup.", - "confidence": 0.87, - "canonical_issue_reason": "Issue 43872 is the strongest canonical bug report: it names a concrete, reproducible bitsandbytes incompatibility and matches the follow-up report about the same unexpected `_is_hf_initialized` constructor argument.", + "summary": "Most items are distinct model- or feature-specific regressions. The only clear duplicate families here are the custom-model initialization regression, the removed docs pipeline-task references, and the repeated mask/tensor shape-mismatch reports.", + "confidence": 0.64, + "canonical_issue_reason": "Issue 43646 is the best umbrella target: it states the broad Transformers 5.0 custom model initialization regression, and 43645 reads like the narrower Jupyter-notebook variant of the same bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43872 is the best overall anchor because it isolates a clear root cause, has direct duplicate evidence, and is more actionable than the broader symptom reports.", + "best_issue_reason": "Issue 43646 is the cleanest representative issue to keep because it is broad, specific enough to identify the regression, and likely subsumes the more specific notebook report.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43827", "right": "issue:44509", "accept": true, - "reason": "Both describe the same v5 documentation cleanup: obsolete pipeline tasks still referenced in docs." + "reason": "Same docs regression: both say removed pipeline task names are still referenced after v5." }, { "left": "issue:43792", "right": "issue:44220", "accept": false, - "reason": "Related Whisper/fbank area, but the titles do not establish the same concrete bug or code path." + "reason": "Related audio/Whisper area, but the titles do not show the same concrete failing code path." }, { "left": "issue:42915", "right": "issue:43872", "accept": false, - "reason": "Different failures: Qwen3Moe FP8 behavior vs a bitsandbytes initialization incompatibility." + "reason": "Different problems: FineGrainedFP8Config vs bitsandbytes Int8Params initialization." }, { "left": "issue:43872", "right": "issue:44291", - "accept": true, - "reason": "Same `_is_hf_initialized`-related constructor regression, just reported through different loading paths." + "accept": false, + "reason": "Same symptom keyword, but different code paths and different load mechanisms; too weak to merge." }, { "left": "issue:42947", "right": "issue:45923", "accept": false, - "reason": "Both touch gradient checkpointing, but they are different models and different failure modes." + "reason": "One is gradient-checkpointing behavior, the other is a missing capability flag; not the same bug." }, { "left": "issue:41553", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading problems, but for different models and with different symptoms." + "reason": "Both are tokenizer-load failures, but for different models and different reported failures." }, { "left": "issue:36331", "right": "issue:41093", "accept": false, - "reason": "Unrelated: custom `compute_loss` signature mismatch vs an index/mask shape error." + "reason": "Unrelated errors: custom trainer signature mismatch vs tensor mask shape mismatch." }, { "left": "issue:43940", "right": "issue:45313", "accept": false, - "reason": "Both are DeepSpeed weight-loading issues, but the affected models and missing-weight patterns differ." + "reason": "Both are ZeRO-3 weight-loading issues, but they target different Qwen model families and likely different fixes." }, { "left": "issue:43257", "right": "issue:45305", "accept": false, - "reason": "Different bugs: missing Qwen3 MOE weight conversion vs incorrect gradient averaging with GAS." + "reason": "Different DeepSpeed/accelerate problems: missing MoE weight conversion vs gradient averaging with loss kwargs." }, { "left": "issue:43854", "right": "issue:44863", "accept": false, - "reason": "Different model families and different load failures." + "reason": "Different models and different loading failures; no clear shared underlying defect." }, { "left": "issue:44928", "right": "issue:45588", "accept": false, - "reason": "Different subsystems and symptoms; no shared concrete code-path bug is evident." + "reason": "Both touch attention/positioning, but one is RLHF gradient instability and the other is a flash-attention AttributeError." }, { "left": "issue:44479", "right": "issue:44560", "accept": false, - "reason": "Both involve video inputs, but they affect different model families and failure modes." + "reason": "Both are video-related Qwen regressions, but the reported failure modes are different." }, { "left": "issue:41093", "right": "issue:44805", - "accept": false, - "reason": "Same generic mask/tensor error text, but not enough evidence that the underlying bug is the same." + "accept": true, + "reason": "Same mask/tensor shape-mismatch error signature; 44805 looks like a later duplicate report of the same bug class." }, { "left": "issue:41553", "right": "issue:44488", "accept": false, - "reason": "Both are model/tokenizer load issues, but they concern different models and likely different root causes." + "reason": "Different tokenizer/model-loading issues with no evidence they share the same code path." }, { "left": "issue:44863", "right": "issue:46402", "accept": false, - "reason": "Both are checkpoint-loading complaints, but they are about different models and different loader behavior." + "reason": "Different models and different loading symptoms; not the same underlying issue." }, { "left": "issue:24643", "right": "issue:30064", "accept": false, - "reason": "Completely different topics: DeepSpeed training error vs segmentation-map preprocessing." + "reason": "Training with DeepSpeed weight-dimension error is unrelated to void segmentation-map preprocessing." }, { "left": "issue:44530", "right": "issue:44863", "accept": false, - "reason": "PagedAttention cache/group-type crash is unrelated to NemotronH checkpoint loading." + "reason": "PagedAttentionCache crash vs checkpoint-loading failure; unrelated." }, { "left": "issue:43366", "right": "issue:43828", "accept": false, - "reason": "Different model support issues with no shared code-path bug." + "reason": "GGUF gpt-oss support and autocast dtype mismatch are unrelated feature/bug reports." }, { "left": "issue:45325", "right": "issue:45902", "accept": false, - "reason": "Different regressions: video rope index scaling vs uninitialized rotary embedding memory." + "reason": "Different Qwen3.x rotary-position bugs with different symptoms and likely different fixes." }, { "left": "issue:45663", "right": "issue:45823", "accept": false, - "reason": "Both involve Gemma4, but one is an FSDP2 attention state bug and the other is a device-map mismatch." + "reason": "Both are Gemma4, but one is a per-layer shared_kv_states KeyError and the other is a device-map mismatch." }, { "left": "issue:43366", "right": "issue:43531", "accept": false, - "reason": "Unrelated model-support vs sliding-window behavior issues." + "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window behavior are not the same issue." }, { "left": "issue:43716", "right": "issue:44560", "accept": false, - "reason": "Different multimedia bugs affecting different model families." + "reason": "Different multimodal regressions: image-preprocessor/model dtype mismatch vs video StopIteration." }, { "left": "issue:43950", "right": "issue:45834", "accept": false, - "reason": "Non-persistent buffer regression and Kosmos2.5 OCR indexing are unrelated." + "reason": "Non-persistent buffer corruption is unrelated to the Kosmos2.5 OCR index error." }, { "left": "issue:43366", "right": "issue:44589", "accept": false, - "reason": "Different model-support issue vs missing Float8 storage type." - }, - { - "left": "issue:44560", - "right": "issue:45685", - "accept": false, - "reason": "Different failures: video StopIteration vs MPS int histogram limitation." + "reason": "Unrelated model-support and storage-type errors." }, { "left": "issue:43645", "right": "issue:43646", "accept": true, - "reason": "These read as the same Transformers 5.0 custom-model initialization regression, with one report adding the Jupyter-specific reproduction." + "reason": "Same underlying custom-model initialization regression; 43645 is just the Jupyter-specific wording of the broader 43646 report." }, { "left": "issue:43550", "right": "issue:44910", "accept": false, - "reason": "Different bugs in different models: SDPA/torch.compile vs flash-attention position_ids handling." + "reason": "Different attention/kernel failures on different models and different symptoms." }, { "left": "issue:45812", "right": "issue:46489", "accept": false, - "reason": "Both tokenizer regressions, but for different model families and likely different fixes." + "reason": "Different tokenizer regressions for different model families and different change histories." }, { "left": "issue:44560", - "right": "issue:44918", - "accept": false, - "reason": "Both are Qwen3.5 video/input issues, but one is a video StopIteration regression and the other is embedding unpacking with TRL." - }, - { - "left": "issue:43854", - "right": "issue:44488", - "accept": false, - "reason": "Different load failures for different model families." - }, - { - "left": "issue:43940", - "right": "issue:45834", - "accept": false, - "reason": "Different problems and different models; no duplicate evidence." - }, - { - "left": "issue:44910", - "right": "issue:44918", + "right": "issue:45685", "accept": false, - "reason": "Both concern Qwen3.5, but one is flash-attention memory access and the other is TRL embedding unpacking." + "reason": "Video model StopIteration and MPS histogram int support are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusions are well grounded in the packet and the soft-edge decisions are appropriately conservative. The accepted pairs are strongly title-aligned, and the rejected pairs avoid overclaiming despite superficial similarity. The canonical/anchor choice on issue 43872 is reasonable given the exact `_is_hf_initialized` constructor error match with 44291." + "feedback": "The summary is grounded in the packet and stays conservative. The accepted soft pairs are supported by the titles and relative specificity (43827/44509, 41093/44805, 43645/43646), and the rejected pairs avoid over-merging on weak symptom similarity. No obvious overstatement." }, "evaluator_used": true, "retried": false, @@ -2020,235 +1936,235 @@ "issue:43257|issue:43854", "issue:45488|issue:45812" ], - "prepared_review_unit_hash": "d3099fe285be55a0d676d4f74ac36cebf51b82116eedf8e51be3d799d5203722", + "prepared_review_unit_hash": "b661dd0f6079909b3a20e65bc27c4cf0f74f999b04344624b2d4639a6fb33ba9", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37088, - "estimated_input_tokens": 9272, - "estimated_eval_tokens": 18800 + "serialized_chars": 37089, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous, with one clear duplicate pair around the v5 non-persistent-buffer junk issue and one likely duplicate SigLIP2 output-regression pair; the remaining items look like distinct bugs or regressions in different code paths.", - "confidence": 0.78, - "canonical_issue_reason": "issue:43644 is the strongest canonical anchor for the non-persistent-buffer corruption reports: it directly describes the v5 junk-buffer symptom, and issue:44534 is the closest near-duplicate with the same bug wording. I would treat this as the best match for that subgroup, not as an exact title-level duplicate.", + "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5/model-specific regressions. Only one soft edge is a clear duplicate; the rest are too cross-model or only share a broad subsystem like loading, quantization, or device mapping.", + "confidence": 0.81, + "canonical_issue_reason": "issue:43644 is the clearest canonical issue: it is earlier, broadly phrased, and has the exact duplicate/rewording in issue:44534.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43644 is the best representative issue in this set because it cleanly captures a concrete, narrowly scoped regression and sits at the center of the closest duplicate pair in the packet.", + "best_issue_reason": "issue:43644 is the best representative of the duplicate set because it captures the shared concrete bug and is the earliest instance.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42915", "right": "issue:43957", "accept": false, - "reason": "Qwen3MoE FP8 config failure and meta-device loading breakage are different problems." + "reason": "Both are loading failures, but one is Qwen3Moe FP8-specific and the other is a generic meta-device loading regression; not the same concrete bug." }, { "left": "issue:43824", "right": "issue:43957", "accept": false, - "reason": "Specific Qwen2.5-VL import failure is unrelated to generic meta-device loading." + "reason": "Import error for a specific class vs a model-loading/meta-device failure; unrelated code paths." }, { "left": "issue:42617", "right": "issue:43366", "accept": false, - "reason": "3d_parallel.py runtime failure and GGUF gpt-oss support are different code paths." + "reason": "3d_parallel.py runtime issue and GGUF/gpt-oss support are different problems." }, { "left": "issue:36010", "right": "issue:42915", "accept": false, - "reason": "GenerationMixin import-path breakage is unrelated to Qwen3MoE FP8 loading." + "reason": "GenerationMixin import failure is unrelated to Qwen3Moe FP8 loading." }, { "left": "issue:44661", "right": "issue:45823", "accept": false, - "reason": "Tokenizer-mapping bug and Gemma4 device-mismatch bug are not the same failure." + "reason": "Tokenizer-mapping bug and Gemma4 device mismatch are different failures." }, { "left": "issue:44560", "right": "issue:44805", "accept": false, - "reason": "Video StopIteration and mask/tensor shape mismatch are distinct errors." + "reason": "Different runtime errors in different model paths; only a broad 'indexing error' resemblance." }, { "left": "issue:38175", "right": "issue:43994", - "accept": true, - "reason": "Both report broken outputs for the same siglip2-base-patch16-224 model and look like the same underlying regression." + "accept": false, + "reason": "Both involve SigLIP2, but the reported symptoms and likely failure modes differ enough to avoid merging without stronger evidence." }, { "left": "issue:43632", "right": "issue:45834", "accept": false, - "reason": "_is_hf_initialized regression is unrelated to Kosmos2.5 long-input index errors." + "reason": "Transformer v5 flag regression and Kosmos2.5 OCR index error are unrelated." }, { "left": "issue:44155", "right": "issue:45584", "accept": false, - "reason": "AudioFlamingo3 track leakage and Whisper empty-transcription failure are separate bugs." + "reason": "AudioFlamingo3 cross-track leakage and Whisper empty-transcription failure are different bugs." }, { "left": "issue:43638", "right": "issue:44661", "accept": false, - "reason": "DeepSpeed Bert index error is not the same as the tokenizer-mapping failure." + "reason": "Bert/DeepSpeed zero3 initialization error is unrelated to the tokenizer-mapping add-new-model-like issue." }, { "left": "issue:43596", "right": "issue:44661", "accept": false, - "reason": "Another Bert ZeRO-3 initialization error; different bug from add-new-model-like failing." + "reason": "Both are load-time failures, but they affect different mechanisms and model families." }, { "left": "issue:44877", "right": "issue:45030", "accept": false, - "reason": "Both are config-loading regressions, but the model targets and failure modes differ too much to merge." + "reason": "Config-validation failures for different models; not the same concrete regression." }, { "left": "issue:44661", "right": "issue:44805", "accept": false, - "reason": "Tokenizer-mapping edge case is unrelated to the mask-index IndexError." + "reason": "Tokenizer-mapping bug and mask/tensor shape mismatch are not the same issue." }, { "left": "issue:43317", "right": "issue:44756", "accept": false, - "reason": "Dequantized offload loading and Strix Halo mmap OOM are different issues." + "reason": "Dequantized offload failure vs mmap/OOM on Strix Halo; too different to treat as duplicates." }, { "left": "issue:38175", "right": "issue:39692", "accept": false, - "reason": "Same model family, but docs-example errors and output-corruption symptoms are not clearly one bug." + "reason": "Same model family, but one is a zero-probability symptom and the other is a docs example with multiple issues, including quantization; not clearly one bug." }, { "left": "issue:42491", "right": "issue:43720", "accept": false, - "reason": "Qwen3 MoE LoRA incompatibility is unrelated to BitNet packed-weight unpacking." + "reason": "Qwen3 MoE LoRA compatibility and BitNet packed-weight loading are distinct code paths." }, { "left": "issue:41720", "right": "issue:42915", "accept": false, - "reason": "CUDA assert under auto device mapping is not the same as the Qwen3MoE FP8 failure." + "reason": "Both are Qwen-related failures, but one is CUDA assert under device mapping and the other is FP8 config loading; not the same bug." }, { "left": "issue:44451", "right": "issue:44488", "accept": false, - "reason": "Both are model-loading failures, but they involve different BERT models and different root causes." + "reason": "Similar 'cannot load model' wording, but different models and insufficient evidence of a single shared root cause." }, { "left": "issue:45313", "right": "issue:45357", "accept": false, - "reason": "ZeRO-3 language_model weight loading and visual-encoder save_pretrained keys are unrelated." + "reason": "ZeRO-3 weight loading and save_pretrained key regression are opposite sides of the workflow and not the same bug." }, { "left": "issue:44960", "right": "issue:44977", "accept": false, - "reason": "GLM5 and Qwen3.5 flash-attention generation failures are distinct regressions." + "reason": "Different model families and different failure modes." }, { "left": "issue:42491", "right": "issue:45276", "accept": false, - "reason": "LoRA reuse on hf5.x and Gemma4 embedding resize behavior are different issues." + "reason": "LoRA/load-compatibility issue vs embedding resize propagation bug; unrelated." }, { "left": "issue:42915", "right": "issue:44756", "accept": false, - "reason": "Qwen3MoE FP8 loading and mmap/OOM behavior are not the same bug." + "reason": "Qwen3Moe FP8 failure and Strix Halo mmap OOM share only a vague memory-loading theme." }, { "left": "issue:43644", "right": "issue:44534", "accept": true, - "reason": "Near-duplicate pair: same v5 non-persistent-buffer corruption symptom and essentially the same report." + "reason": "Same bug: identical title/body theme about non-persistent buffers being filled with junk in Transformers v5." }, { "left": "issue:45507", "right": "issue:46468", "accept": false, - "reason": "GraniteMoEHybrid invalid method call and Nemotron example breakage do not share a concrete code path." + "reason": "GraniteMoEHybrid invalid-method calls and Nemotron example breakage are different issues." }, { "left": "issue:39692", "right": "issue:43994", "accept": false, - "reason": "Docs-example/model-processor mismatch and general bad inference output are related only at a very broad model-family level." + "reason": "The documentation example bug and the nonsensical-output report are related at best by model family, but not clearly the same concrete defect." }, { "left": "issue:43708", "right": "issue:44075", "accept": false, - "reason": "Checkpoint-resume max_steps bug is unrelated to SGD argument handling." + "reason": "Trainer checkpoint math and SGD argument handling are different trainer bugs." }, { "left": "issue:43856", "right": "issue:43873", "accept": false, - "reason": "Training memory inefficiency and quantization offloading behavior are not the same issue." + "reason": "Both mention quantization/offloading/memory, but one is training memory inefficiency and the other is offload not working as expected." }, { "left": "issue:43644", "right": "issue:43645", "accept": false, - "reason": "Buffer initialization regression and Jupyter custom-model initialization breakage are different bugs." + "reason": "Both are Transformers v5 regressions, but one is buffer initialization and the other is custom-model initialization in notebooks; not the same bug." }, { "left": "issue:43854", "right": "issue:43931", "accept": false, - "reason": "GLM-4.7-Flash unit-test loading failure and Qwen3-VL shape mismatch are separate model-specific problems." + "reason": "Different models and different loading errors." }, { "left": "issue:43856", "right": "issue:44387", "accept": false, - "reason": "General memory inefficiency and int4 reserved-memory OOM are related only broadly, not one concrete bug." + "reason": "Both concern memory, but one is Qwen3 MoE training efficiency and the other is int4 reserved-memory OOM under Transformers 5.x." }, { "left": "issue:43257", "right": "issue:43854", "accept": false, - "reason": "Qwen3 MOE accelerate/deepspeed conversion and GLM-4.7-Flash unit-test loading are unrelated." + "reason": "Qwen3 MoE weight conversion/loading and GLM-4.7-Flash unit-test loading are unrelated." }, { "left": "issue:45488", "right": "issue:45812", "accept": false, - "reason": "Tokenizer regression and Granite AutoTokenizer regression hit different tokenizer paths and model families." + "reason": "Different tokenizer regressions affecting different model families; same broad area, not the same concrete bug." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The cluster summary is mostly grounded, but the canonical anchor rationale overstates the evidence. `issue:43644` and `issue:44534` are very close matches, but not an exact title-level duplicate, and the claim that 43644 has the \"clearer discussion trail\" is not supported by the packet (discussion activity is comparable). Please soften that to a near-duplicate / strongest match claim rather than asserting exact duplication or stronger evidence than is present." + "accept": true, + "feedback": "The lone positive edge (issue:43644 \u2194 issue:44534) is well grounded by near-identical wording and the earlier issue is a reasonable canonical representative. The remaining soft edges are phrased conservatively, generally distinguishing model-specific or code-path-specific failures without overclaiming duplication. Overall, the summary stays within the evidence in the packet." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -2351,20 +2267,20 @@ "issue:45588|issue:45663", "issue:41720|issue:44155" ], - "prepared_review_unit_hash": "501e59994b0f8359ba04983773d45edf260e1187ee9b4bd146f445cd1abc19ed", + "prepared_review_unit_hash": "93f370129fe54111b069fe41aaff631084478123fc53f675f4527f1995327105", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 34, - "serialized_chars": 37962, + "serialized_chars": 37964, "estimated_input_tokens": 9491, "estimated_eval_tokens": 19238 }, @@ -2373,220 +2289,223 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated Transformers bug reports; only the resize_token_embeddings/output_embeddings regression pair looks like a true duplicate cluster.", - "confidence": 0.79, - "canonical_issue_reason": "issue:45292 is the broader, model-agnostic report of the resize_token_embeddings/output_embeddings regression, while issue:45276 appears to be a Gemma4-specific manifestation of the same bug.", + "summary": "This cluster is a loose collection of mostly unrelated Transformers bugs. A few are about save/load or token embedding propagation, but the soft pairs do not look like true duplicates; they mostly share a subsystem, not the same underlying defect.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45698 is the best single representative because it is the broadest active save/load regression in the set and is closest in theme to several neighboring custom-module and save_pretrained/from_pretrained reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45292 is the best representative issue because it states the underlying failure in the most general form and covers the same bug as issue:45276.", + "best_issue_reason": "Issue 45698 is the most representative cluster anchor: it is general, recent, open, and centered on save/load behavior rather than a narrow model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:42491", "right": "issue:45292", "accept": false, - "reason": "Different problems: qwen3_moe LoRA compatibility/loading vs a resize_token_embeddings/output_embeddings issue." + "reason": "Both involve model/config loading and embedding-related behavior, but one is a Qwen3 MoE LoRA compatibility problem and the other is a resize_token_embeddings propagation bug." }, { "left": "issue:42491", "right": "issue:43931", "accept": false, - "reason": "Different model families and failure modes; one is LoRA checkpoint incompatibility, the other is a Qwen3-VL weight-shape mismatch on load." + "reason": "Different models and failures: Qwen3 MoE LoRA incompatibility vs Qwen3-VL weight-shape mismatch." }, { "left": "issue:44488", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer/model-loading failures, but the affected models and likely root causes are distinct; not enough evidence of the same underlying bug." + "reason": "Both are tokenizer/model load failures, but they affect different models and describe different concrete breakages." }, { "left": "issue:45276", "right": "issue:45335", - "accept": true, - "reason": "Same resize_token_embeddings bug: one generic report about output_embeddings, the other a T5Gemma-specific variant affecting decoder.embed_tokens." + "accept": false, + "reason": "Same general API area, but Gemma4-specific missing propagation to multiple embeddings is not clearly the same bug as the T5Gemma decoder embed_tokens issue." }, { "left": "issue:44910", "right": "issue:45588", "accept": false, - "reason": "Both involve flash_attention, but they describe different crashes and different triggers/code paths." + "reason": "Both are flash-attention related crashes, but the Qwen3.5 position_ids bug and the sink-less s_aux=None AttributeError are different code paths." }, { "left": "issue:44164", "right": "issue:44291", "accept": false, - "reason": "save/from_pretrained extra_state handling is separate from the init_empty_weights _is_hf_initialized argument regression." + "reason": "Both touch model loading/saving, but extra_state handling and init_empty_weights/_is_hf_initialized are separate bugs." }, { "left": "issue:45362", "right": "issue:45406", "accept": false, - "reason": "Different components and failures: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer in transformers serve." + "reason": "Different surfaces and failures: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer in transformers serve." }, { "left": "issue:43404", "right": "issue:43454", "accept": false, - "reason": "Both mention untied lm_head weights, but they are different model implementations and no shared concrete code-path is evident." + "reason": "Both mention untied lm_head, but they are different model implementations and likely separate missing-tie bugs." }, { "left": "issue:43425", "right": "issue:43957", "accept": false, - "reason": "Unrelated issues: Torch 2.10 compatibility vs meta-device model loading breakage." + "reason": "Torch 2.10 compatibility and meta-device loading failures are not the same underlying issue." }, { "left": "issue:39692", "right": "issue:44960", "accept": false, - "reason": "Documentation example errors for SigLIP2 are unrelated to the GLM5 issue." + "reason": "A broken documentation example and a GLM5 loading issue are unrelated." }, { "left": "issue:44402", "right": "issue:44488", "accept": false, - "reason": "Different token-vocab/loading problems for different models; not the same bug." + "reason": "Both are tokenizer-related, but one is vocab size mismatch and the other is a model load failure for a different checkpoint." }, { "left": "issue:42947", "right": "issue:45704", "accept": false, - "reason": "Gradient checkpointing/LoRA behavior is unrelated to the T5 apex.FusedRMSNorm memory leak." + "reason": "Gradient checkpointing/LoRA and an apex FusedRMSNorm memory leak are unrelated bugs." }, { "left": "issue:36010", "right": "issue:39404", "accept": false, - "reason": "ImportError for GenerationMixin is unrelated to Whisper return_language pipeline regression." + "reason": "ImportError for GenerationMixin and Whisper pipeline return_language regression are unrelated." }, { "left": "issue:45663", "right": "issue:45759", "accept": false, - "reason": "Gemma4 FSDP2 shared_kv_states issue is unrelated to Qwen multimodal text_config unwrapping." + "reason": "Gemma4 FSDP2 shared_kv_states KeyError and Qwen composite-config unwrapping are different model-specific issues." }, { "left": "issue:43873", "right": "issue:44387", "accept": false, - "reason": "Both concern quantization/offloading memory behavior, but the specific symptoms and code paths differ." + "reason": "Both involve memory behavior under quantization, but one is generic offloading and the other is int4 reserved-memory/OOM growth." }, { "left": "issue:45663", "right": "issue:45859", "accept": false, - "reason": "Gemma4 FSDP2 attention state bug is unrelated to Qwen3_5Moe rotary embedding CPU offload." + "reason": "Gemma4 training FSDP2 bug and Qwen3_5Moe CPU-offload incompatibility are different failures." }, { "left": "issue:44155", "right": "issue:45823", "accept": false, - "reason": "AudioFlamingo3 batched inference leakage is unrelated to Gemma4 PLE device mismatch." + "reason": "AudioFlamingo3 batched inference leakage is unrelated to Gemma4 device-map mismatch." }, { "left": "issue:44291", "right": "issue:44756", "accept": false, - "reason": "Different loading bugs: init_empty_weights argument handling vs mmap/OOM on Strix Halo." + "reason": "Unexpected init_empty_weights argument and Strix Halo mmap/OOM are unrelated." }, { "left": "issue:44292", "right": "issue:44912", "accept": false, - "reason": "Model loading type/shape mismatch is different from the MXFP4 quantization fallback problem." + "reason": "init_empty_weights/_is_hf_initialized loading failure is unrelated to MXFP4 quantization fallback." }, { "left": "issue:44960", "right": "issue:46050", "accept": false, - "reason": "GLM5 model-loading issue is unrelated to sinq save_pretrained/from_pretrained quantization persistence." + "reason": "GLM5 loading and sinq save/load corruption are different model-specific issues." }, { "left": "issue:44387", "right": "issue:44756", "accept": false, - "reason": "Different resource issues and triggers; int4 CUDA reserved-memory growth is not the same as mmap OOM on Strix Halo." + "reason": "Both mention OOM, but the causes differ: int4 reserved-memory growth vs mmap on Strix Halo." }, { "left": "issue:43335", "right": "issue:45663", "accept": false, - "reason": "SwitchTransformers sparse layer construction bug is unrelated to Gemma4 FSDP2 attention state rebuilding." + "reason": "Sparse-layer creation in SwitchTransformers and Gemma4 FSDP2 shared_kv_states are unrelated." }, { "left": "issue:42947", "right": "issue:43381", "accept": false, - "reason": "Both mention gradient checkpointing, but one is PEFT LoRA ineffectiveness and the other is eval-mode incompatibility." + "reason": "Both are gradient-checkpointing related, but one is ineffective LoRA training and the other is an eval-mode restriction." }, { "left": "issue:45276", "right": "issue:45292", - "accept": true, - "reason": "These describe the same resize_token_embeddings regression; 45276 is the Gemma4-specific instance and 45292 is the generic report." + "accept": false, + "reason": "Highly similar theme, but the Gemma4-specific multi-embedding propagation bug is not clearly the same defect as the general output_embeddings issue." }, { "left": "issue:44863", "right": "issue:46468", "accept": false, - "reason": "NemotronH checkpoint loading and the official Nemotron example failure are related by model family but not clearly the same concrete bug." + "reason": "NemotronH checkpoint loading and the official Nemotron example failure are related by model family only, not the same bug." }, { "left": "issue:45685", "right": "issue:46468", "accept": false, - "reason": "MPS histogram_mps integer support is unrelated to the Nemotron example loading issue." + "reason": "An MPS histogram kernel limitation is unrelated to the Nemotron example loading breakage." }, { "left": "issue:43854", "right": "issue:45216", "accept": false, - "reason": "Different regressions: GLM-4.7-Flash unit-test loading vs Qwen3.5 save_pretrained checkpoint correctness." + "reason": "GLM-4.7-Flash unit-test loading and Qwen3.5 save_pretrained regression are separate model-specific bugs." }, { "left": "issue:45684", "right": "issue:45698", "accept": false, - "reason": "Both involve custom-model save/load, but one is a permissions propagation bug and the other is wrong-module loading after save_pretrained." + "reason": "Both are save_pretrained-related, but one is file permission propagation and the other is wrong custom-module loading." }, { "left": "issue:44479", "right": "issue:44877", "accept": false, - "reason": "Video-input regression across Qwen models is unrelated to strict Granite Speech config loading." + "reason": "Video-input regression and strict Granite Speech config loading are different issues." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction and Qwen3.5 RLHF NaN explosion are not the same bug." }, { "left": "issue:44164", "right": "issue:44756", "accept": false, - "reason": "extra_state save/load handling is unrelated to mmap disabling on Strix Halo." + "reason": "save/from_pretrained extra_state handling is unrelated to mmap-induced OOM on Strix Halo." }, { "left": "issue:43957", "right": "issue:44756", "accept": false, - "reason": "Meta-device loading breakage is unrelated to the Strix Halo mmap/OOM issue." + "reason": "Meta-device loading failures and Strix Halo mmap OOM are different code paths." }, { "left": "issue:45588", "right": "issue:45663", "accept": false, - "reason": "Flash-attention sink-less-model AttributeError is unrelated to Gemma4 FSDP2 shared_kv_states rebuilding." + "reason": "Flash-attention sink-less-model crash and Gemma4 FSDP2 KeyError are unrelated." }, { "left": "issue:41720", "right": "issue:44155", "accept": false, - "reason": "Qwen3 auto device mapping cudaErrorAssert is unrelated to AudioFlamingo3 batched inference leakage." + "reason": "Qwen3 auto device-map cuda assert and AudioFlamingo3 batched track leakage are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The clustering is conservative and grounded in the packet. The only accepted links are the resize_token_embeddings regressions, and the rest are reasonably rejected as different model families, symptoms, or code paths. The canonical issue choice (45292 over 45276) is also well supported because 45292 is the broader formulation of the same embedding-update bug." - }, - "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 }, @@ -2676,8 +2595,8 @@ "issue:43901|issue:44509", "issue:43425|issue:44918", "issue:39404|issue:44912", - "issue:36032|issue:43452", "issue:43596|issue:44683", + "issue:36032|issue:43452", "issue:44164|issue:45759", "issue:43299|issue:43931", "issue:44991|issue:45081", @@ -2686,232 +2605,232 @@ "issue:43856|issue:44756", "issue:42796|issue:43404" ], - "prepared_review_unit_hash": "c981e3dc62a3c443c03a0679312ed288894b3fde575d5808bbba5c883bbc9998", + "prepared_review_unit_hash": "789374209ffa92eb8154344446f1914253b1c7feeca1d6ae435f4307e4ce88f4", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37479, - "estimated_input_tokens": 9370, - "estimated_eval_tokens": 18996 + "serialized_chars": 37476, + "estimated_input_tokens": 9369, + "estimated_eval_tokens": 18994 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issues. The only convincing duplicate pair is the resize_token_embeddings regression: a generic report and a T5Gemma-specific manifestation.", - "confidence": 0.79, - "canonical_issue_reason": "issue:45292 is the broader, model-agnostic report about resize_token_embeddings not updating output embeddings; issue:45335 looks like a specific instance of that same bug in t5gemma.", + "summary": "This cluster is mostly heterogeneous regression reports around tokenizers/model loading, with one clear duplicate pair about `resize_token_embeddings` not propagating to all tied output embeddings. I did not find any PRs here.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45292 is the broadest statement of the `resize_token_embeddings` propagation bug; issue:45335 is a model-specific instance of the same missing update to downstream embeddings.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45292 is the best canonical issue because it describes the underlying bug most generally and can subsume the T5Gemma-specific report.", + "best_issue_reason": "issue:45292 best represents the underlying bug because it describes the general embedding-resize failure, rather than a single model variant.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43646", "right": "issue:43708", "accept": false, - "reason": "Different bugs: custom model init in notebooks vs Trainer resume/max_steps calculation." + "reason": "Different failures: custom model init/loading vs Trainer checkpoint resume step counting." }, { "left": "issue:45292", "right": "issue:45335", "accept": true, - "reason": "Same underlying regression: resize_token_embeddings fails to propagate to tied/output embeddings; 45335 is a model-specific manifestation." + "reason": "Same underlying bug: `resize_token_embeddings` updates the main embeddings but not the tied/output embeddings; 45335 is a t5gemma-specific instance." }, { "left": "issue:43645", "right": "issue:44164", "accept": false, - "reason": "Different code paths: notebook custom model initialization vs save/from_pretrained extra_state handling." + "reason": "Notebook custom-model init regression is unrelated to `extra_state` save/load handling." }, { "left": "issue:43828", "right": "issue:43957", "accept": false, - "reason": "Both are mismatch/loading problems, but one is autocast dtype handling and the other is meta-device model loading." + "reason": "Both are dtype-related, but one is autocast inference and the other is meta-device loading; different code paths." }, { "left": "issue:46315", "right": "issue:46491", "accept": false, - "reason": "Tokenizer performance regression vs leading-space round-trip correctness bug; not the same issue." + "reason": "Different tokenizer regressions: slow `convert_tokens_to_ids` performance vs CodeLlama round-trip spacing." }, { "left": "issue:44402", "right": "issue:44451", "accept": false, - "reason": "One is vocab-size/token placement, the other is a specific model loading failure; insufficient evidence of the same root cause." + "reason": "A specific tokenizer vocab/special-token issue vs a generic model loading failure; not clearly the same bug." }, { "left": "issue:43425", "right": "issue:43828", "accept": false, - "reason": "General torch compatibility complaint vs a specific autocast dtype mismatch." + "reason": "Torch compatibility report vs autocast dtype mismatch; unrelated problems." }, { "left": "issue:45812", "right": "issue:46315", "accept": false, - "reason": "Wrong token IDs for Granite models is a correctness regression; convert_tokens_to_ids slowdown is a performance issue." + "reason": "Granite token ID regression and slow `convert_tokens_to_ids` performance are different tokenizer bugs." }, { "left": "issue:44589", "right": "issue:44756", "accept": false, - "reason": "Float8 storage lookup failure and Strix Halo mmap OOM are unrelated." + "reason": "Float8 storage lookup failure vs mmap/OOM on Strix Halo; different symptoms and causes." }, { "left": "issue:44683", "right": "issue:45823", "accept": false, - "reason": "Compiled flex_attention failure on torch >=2.9 is unrelated to Gemma4 device-mismatch during forward." + "reason": "Compiled flex_attention on torch>=2.9 is unrelated to Gemma4 device-mismatch during forward." }, { "left": "issue:43611", "right": "issue:43645", "accept": false, - "reason": "Base_model_prefix loading breakage and custom model init in notebooks are distinct failures." + "reason": "Both involve model loading, but one is `base_model_prefix` handling and the other is custom model initialization in notebooks." }, { "left": "issue:43425", "right": "issue:45823", "accept": false, - "reason": "Torch version incompatibility and Gemma4 PLE device mismatch are different problems." + "reason": "Torch 2.10 incompatibility is not the same as a Gemma4 device-map forward mismatch." }, { "left": "issue:43425", "right": "issue:43716", "accept": false, - "reason": "General torch compatibility issue vs Mistral-3 image preprocessor/model dtype mismatch." + "reason": "General torch incompatibility vs a specific image-preprocessor/model dtype mismatch." }, { "left": "issue:44530", "right": "issue:45084", "accept": false, - "reason": "Different failures: Qwen3.5 PagedAttentionCache group type vs template compilation error." + "reason": "PagedAttentionCache linear_attention crash and non-template-node compilation error are different issues." }, { "left": "issue:39404", "right": "issue:45701", "accept": false, - "reason": "Whisper pipeline return_language regression is unrelated to tokenization changes across versions." + "reason": "Whisper `return_language` pipeline behavior is not the same as a broad tokenization change report." }, { "left": "issue:42796", "right": "issue:46326", "accept": false, - "reason": "Mistral tokenizer BPE marker handling and continue_final_message behavior are distinct tokenizer issues." + "reason": "Ministral BPE-marker handling and `continue_final_message` for mistral_common are related to Mistral tokenization but not the same concrete bug." }, { "left": "issue:43906", "right": "issue:44530", "accept": false, - "reason": "An isolated reproduction of an old issue is not the same as the Qwen3.5 PagedAttentionCache crash." + "reason": "Isolated reproduction of a prior issue vs a Qwen3.5 PagedAttentionCache crash; no direct match." }, { "left": "issue:44265", "right": "issue:44898", "accept": false, - "reason": "torch.export / torch_compilable_check failure is unrelated to Perceiver interpolation handling." + "reason": "torch.export compatibility and Perceiver interpolation failure are unrelated model/export bugs." }, { "left": "issue:44589", "right": "issue:45084", "accept": false, - "reason": "Different root causes: missing Float8 storage object vs compilation/template-node error." + "reason": "Different failure modes: missing Float8 storage object vs template-node compilation error." }, { "left": "issue:44479", "right": "issue:44805", "accept": false, - "reason": "Video-input regression for Qwen VL models and mask-shape IndexError are not the same bug." + "reason": "Video-input regression for Qwen VL models vs a mask shape IndexError; not the same bug." }, { "left": "issue:43901", "right": "issue:44509", "accept": false, - "reason": "Both are docs-related, but they cover different removed/changed pipeline docs and are not a single duplicate." + "reason": "Both are docs regressions, but they concern different removed/changed pipeline APIs." }, { "left": "issue:43425", "right": "issue:44918", "accept": false, - "reason": "Torch compatibility complaint vs TRL unpacking failure for Qwen3.5 embeddings." + "reason": "Torch incompatibility is unrelated to TRL SFT unpacking of Qwen3.5 embeddings." }, { "left": "issue:39404", "right": "issue:44912", "accept": false, - "reason": "Whisper return_language regression is unrelated to MXFP4 quantization fallback." + "reason": "Whisper pipeline language return and MXFP4 quantization fallback are different regressions." }, { - "left": "issue:36032", - "right": "issue:43452", + "left": "issue:43596", + "right": "issue:44683", "accept": false, - "reason": "Tokenizer attribute conflict and gguf_file/from_pretrained breakage are unrelated." + "reason": "Deepspeed zero3/BertModel index error is unrelated to flex_attention compilation on newer torch." }, { - "left": "issue:43596", - "right": "issue:44683", + "left": "issue:36032", + "right": "issue:43452", "accept": false, - "reason": "Zero3/BertModel index error and compiled flex_attention failure are different code paths." + "reason": "T5 tokenizer load conflict and `gguf_file` loading breakage are different tokenizer/model-loading bugs." }, { "left": "issue:44164", "right": "issue:45759", "accept": false, - "reason": "extra_state save/load handling and text_config unwrapping are separate model-configuration bugs." + "reason": "`extra_state` serialization and composite Qwen config unwrapping are separate loading paths." }, { "left": "issue:43299", "right": "issue:43931", "accept": false, - "reason": "Both touch Qwen3-VL loading, but one is a MoE loading regression and the other is a specific weight-shape mismatch; not clearly the same fix." + "reason": "Same model family, but one is a broad v5 dev loading regression and the other is a specific weight-shape mismatch; not clearly the same concrete bug." }, { "left": "issue:44991", "right": "issue:45081", "accept": false, - "reason": "Tokenizer load failure for EMBEDDIA/est-roberta and Mistral regex patch crash are different tokenizer bugs." + "reason": "Tokenizer load failure for est-roberta and Mistral regex patch crash are unrelated." }, { "left": "issue:44671", "right": "issue:44991", "accept": false, - "reason": "CamemBERT masked-LM prediction regression is unrelated to tokenizer loading failure for est-roberta." + "reason": "CamemBERT masked-LM prediction regression and tokenizer loading failure for a different model are distinct issues." }, { "left": "issue:43994", "right": "issue:44079", "accept": false, - "reason": "Siglip2 nonsensical outputs and ModelOutput key assignment are unrelated." + "reason": "Siglip2 wrong outputs and `ModelOutput` key assignment are not the same code-path problem." }, { "left": "issue:43856", "right": "issue:44756", "accept": false, - "reason": "Qwen3 MoE memory inefficiency and Strix Halo mmap OOM are different issues." + "reason": "Qwen3 MoE training memory usage and disabling mmap for Strix Halo OOM are unrelated." }, { "left": "issue:42796", "right": "issue:43404", "accept": false, - "reason": "Tokenizer BPE-marker handling and lm_head weight tying are unrelated." + "reason": "Tokenizer BPE-marker handling is unrelated to Mistral3 lm_head weight tying." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary stays within the evidence in the packet. The only accepted soft pair, issue:45292 vs issue:45335, is a reasonable conservative duplicate call based on their closely aligned titles and bug descriptions. The rest of the soft-edge rejections are also cautious and do not overclaim shared root causes." + "feedback": "The summary is grounded in the packet, and the only accepted soft pair (issue:45292 / issue:45335) is a conservative duplicate-style match on the same resize-token-embeddings propagation bug. The rest of the soft-edge verdicts stay appropriately cautious, and it\u2019s correct that no PRs were present." }, "evaluator_used": true, "retried": false, @@ -3013,221 +2932,224 @@ "issue:45084|issue:45507", "issue:42175|issue:46178" ], - "prepared_review_unit_hash": "8bfa012c3b48f762af8598354fbfd279f1a22bdc7f3454c0db4d667425a00e79", + "prepared_review_unit_hash": "c1bd690c3337cce6ad5d270afc45f1c4d7da197ec3bef47e9a51abcd5fe2d544", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36653, - "estimated_input_tokens": 9164, - "estimated_eval_tokens": 18584 + "serialized_chars": 36648, + "estimated_input_tokens": 9162, + "estimated_eval_tokens": 18580 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose set of Transformers regressions centered on v5/v5.1+ loading, tokenizer, model-init, and audio/Whisper issues, but the soft pairs mostly look like neighboring reports rather than true duplicates. I would keep the broad custom-model-init regression as the representative issue.", - "confidence": 0.56, - "canonical_issue_reason": "issue:43646 is the broadest representative title in the set: it captures the general v5 custom-model initialization regression that several nearby reports are loosely related to.", + "summary": "The set is mostly a grab bag of unrelated issues; only one soft duplicate looks plausible. The strongest anchor is the generic GGUF-loading regression, with the model-specific Qwen2.5 report as its likely duplicate.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43452 is the broadest and most generic report of the GGUF-loading failure, so it best serves as the anchor for the only plausible duplicate pair.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43646 is the best single issue to anchor the cluster because it is broad, clearly regression-focused, and more representative than the narrower buffer-specific or model-specific reports.", + "best_issue_reason": "issue:43452 is the clearest representative of the shared GGUF loader regression and is better suited as the canonical issue than the model-specific follow-up.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44462", "right": "issue:46489", "accept": false, - "reason": "Both are tokenizer-loading regressions, but 44462 is about ignoring tokenizer.json while 46489 is a DeepSeek-Coder v1 output regression tied to a specific fix gap; not the same bug." + "reason": "Both are tokenizer regressions, but one is about repository tokenizer.json resolution while the other is a DeepSeek-Coder-specific output bug; not clearly the same code path." }, { "left": "issue:44164", "right": "issue:44479", "accept": false, - "reason": "Completely different areas: save/from_pretrained extra_state handling vs a Qwen video-input regression." + "reason": "Unrelated failures: save/from_pretrained extra_state handling vs a video-input regression in Qwen VL models." }, { "left": "issue:43906", "right": "issue:46326", "accept": false, - "reason": "Both touch tokenizer/mistral_common behavior, but one is an isolated reproduction of an older issue and the other is a specific continue_final_message bug; insufficient evidence of the same underlying defect." + "reason": "Both mention tokenizer-related behavior, but one is an isolated repro of another issue and the other is a mistral_common continue_final_message bug; not the same defect." }, { "left": "issue:44737", "right": "issue:45859", "accept": false, - "reason": "Both are device-related performance/compatibility problems, but they affect different models and code paths (XLNet positional encoding vs Qwen3.5 MoE offload)." + "reason": "Both involve device/offload behavior, but they affect different models and different code paths." }, { "left": "issue:44661", "right": "issue:44683", "accept": false, - "reason": "Unrelated failures: tokenizer/model registration logic vs compiled flex_attention on newer torch." + "reason": "Different subsystems entirely: tokenizer mapping discovery vs torch 2.9 flex_attention compilation." }, { "left": "issue:44488", "right": "issue:45081", "accept": false, - "reason": "Both are tokenizer loading issues, but one is a model load failure and the other is a specific Mistral regex patch crash; different concrete bugs." + "reason": "Both are tokenizer-loading problems, but one is model-loading for a BERT repo and the other is a Mistral regex patch crash." }, { "left": "issue:39404", "right": "issue:44671", "accept": false, - "reason": "Whisper pipeline return_language regression and CamemBERT masked-LM output regression are unrelated." + "reason": "Whisper return_language pipeline regression is unrelated to CamemBERT masked-LM prediction changes." }, { "left": "issue:43611", "right": "issue:43644", "accept": false, - "reason": "Both are v5 regressions, but one is about base_model_prefix loading and the other about non-persistent buffers being initialized with junk; not the same code-path problem." + "reason": "Both are v5 regressions, but one is base_model_prefix loading and the other is non-persistent buffer initialization; different bugs." }, { "left": "issue:46285", "right": "issue:46402", "accept": false, - "reason": "Different model families and symptoms: CLIP model structure inconsistency vs Sam3Model text encoder weights not loading." + "reason": "Different models and symptoms: CLIP structure drift vs Sam3Model text encoder weights not loading." }, { "left": "issue:43716", "right": "issue:45084", "accept": false, - "reason": "Different subsystems entirely: multimodal dtype mismatch vs a template-compilation error." + "reason": "Unrelated: image-preprocessor dtype mismatch vs template compilation error." }, { "left": "issue:43582", "right": "issue:43632", "accept": false, - "reason": "Apple Silicon allocator warmup TypeError is unrelated to the _is_hf_initialized flag regression." + "reason": "Different regression classes: Apple Silicon allocator warmup TypeError vs _is_hf_initialized flag breakage." }, { "left": "issue:40990", "right": "issue:43421", "accept": false, - "reason": "Perplexity/model-quality issue vs tokenizer backend post-processor update bug; not the same defect." + "reason": "Perplexity issue on GPT-OSS is unrelated to tokenizer backend post-processor updates." }, { "left": "issue:44488", "right": "issue:44987", "accept": false, - "reason": "Both are model-loading complaints, but they target different repositories/models and do not clearly share a single failing code path." + "reason": "Both are load failures, but for different models and likely different root causes." }, { "left": "issue:32090", "right": "issue:39404", "accept": false, - "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper return_language behavior." + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper pipeline language-return behavior." }, { "left": "issue:36032", "right": "issue:42915", "accept": false, - "reason": "Tokenizer special-token conflict in T5 is unrelated to Qwen3Moe FineGrainedFP8Config failure." + "reason": "Tokenizer special-token method conflict vs Qwen3Moe FP8 configuration failure are unrelated." }, { "left": "issue:43643", "right": "issue:44530", "accept": false, - "reason": "AutoConfig trust_remote_code field loss is not the same issue as PagedAttentionCache crashing on linear_attention groups." + "reason": "AutoConfig trust_remote_code field loss is unrelated to Qwen3.5 cache/group-type crashes." }, { "left": "issue:46050", "right": "issue:46326", "accept": false, - "reason": "Quantized save/load breakage and mistral_common continue_final_message are unrelated." + "reason": "Quantized model save/load breakage is unrelated to the mistral_common continue_final_message issue." }, { "left": "issue:34567", "right": "issue:40444", "accept": false, - "reason": "TrainerState token counting bug is unrelated to finetuning with multi-image IterableDataset failures." + "reason": "TrainerState token counting and Qwen2.5-VL iterable dataset failures are different bugs." }, { "left": "issue:44451", "right": "issue:44987", "accept": false, - "reason": "Both concern loading models, but they are different model-specific failures with no clear shared bug." + "reason": "Different model-loading regressions affecting unrelated repositories." }, { "left": "issue:43541", "right": "issue:45084", "accept": false, - "reason": "Both mention tracing/compilation, but one is a grouped_mm Mixtral tracing failure and the other is a template-node compilation error." + "reason": "Torch Dynamo grouped_mm tracing error is not the same as non-template-node compilation failure." }, { "left": "issue:44811", "right": "issue:45584", "accept": false, - "reason": "Both are Whisper-related, but one is batch_decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens." + "reason": "Both are Whisper-related, but one is decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens." }, { "left": "issue:42915", "right": "issue:44530", "accept": false, - "reason": "Different Qwen issues with different subsystems: FP8 config vs cache/group-type crash." + "reason": "Both are Qwen-family issues, but one is FP8 config setup and the other is paged attention cache/group-type handling." }, { "left": "issue:43644", "right": "issue:43646", "accept": false, - "reason": "Both are v5 custom-model regressions, but 43644 is specifically about junk-filled non-persistent buffers while 43646 is the broader custom-model initialization breakage; they are related but not the same bug." + "reason": "Different v5 regressions: junk buffers vs custom model initialization." }, { "left": "issue:43452", "right": "issue:43482", - "accept": false, - "reason": "Both involve GGUF loading, but 43452 is a general AutoTokenizer/AutoModelForCausalLM from_pretrained breakage and 43482 is a Qwen2.5-GGUF-specific v5 failure; not clearly the same underlying defect." + "accept": true, + "reason": "Both point to the same GGUF loading path breaking in transformers v5; 43482 looks like a model-specific manifestation of the broader 43452 regression." }, { "left": "issue:43883", "right": "issue:44488", "accept": false, - "reason": "Molmo missing all_tied_weights_keys is unrelated to loading cjvt/sleng-bert." + "reason": "Molmo tied-weights attribute error is unrelated to model-load failures for BERT repositories." }, { "left": "issue:44206", "right": "issue:44220", "accept": false, - "reason": "Both touch audio feature extraction, but 44206 is specifically about an unsupported center arg in LasrFeatureExtractor while 44220 is a generic _torch_extract_fbank_features problem; too little evidence they are the same bug." + "reason": "Related audio feature extraction area, but one is a v5.2.0 regression from an unsupported center arg and the other is a different extract_fbank issue." }, { "left": "issue:43317", "right": "issue:43482", "accept": false, - "reason": "Device-map/offload dequantization failure is unrelated to Qwen2.5 GGUF loading." + "reason": "Dequantized-model device_map offload failure is not the same as a GGUF loading regression." }, { "left": "issue:36032", "right": "issue:43645", "accept": false, - "reason": "T5 tokenizer special-token conflict and Jupyter custom-model initialization are different regressions." + "reason": "Tokenizer add_special_tokens conflict and Jupyter custom model initialization are unrelated." }, { "left": "issue:45084", "right": "issue:45507", "accept": false, - "reason": "Template compilation failure and GraniteMoEHybrid invalid-method call are different concrete problems." + "reason": "Different problems: template compilation error vs GraniteMoEHybrid calling an invalid method." }, { "left": "issue:42175", "right": "issue:46178", "accept": false, - "reason": "One is missing TensorFlow in the torch extra install, the other is a backend mapping validation failure for tensorflow_text; related theme, different bug." + "reason": "TensorFlow backend packaging issue is unrelated to tensorflow_text backend mapping validation." } ] }, - "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\u2014issue:43452 with issue:43482\u2014is reasonably supported by the titles and the broad GGUF-loading regression framing. The rest of the negative verdicts are cautious and do not overclaim overlap." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -3325,226 +3247,226 @@ "issue:43425|issue:44403", "issue:43296|issue:43531" ], - "prepared_review_unit_hash": "288d4a058700f4ea8c65ba60ac404dfa8a5432419514f530c71205cbf518cc6f", + "prepared_review_unit_hash": "9aa1dc935ab9cc153569fe90d0c8bee70063c940b0474d055e9595767269f25d", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 31, - "serialized_chars": 36524, - "estimated_input_tokens": 9131, - "estimated_eval_tokens": 18518 + "serialized_chars": 36517, + "estimated_input_tokens": 9130, + "estimated_eval_tokens": 18516 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly a grab bag of unrelated Transformers regressions and environment-specific failures. The only plausible duplicate/follow-up pairs are the custom-model init reports and the Qwen3.5 save_pretrained reports; the rest target different code paths, models, or failure modes.", - "confidence": 0.88, - "canonical_issue_reason": "issue:45812 is the best representative issue: it describes a broad, concrete v4\u2192v5 tokenizer regression affecting all Granite models, with clear symptoms and wide impact.", + "summary": "Most items are distinct bugs/regressions across different models or subsystems. The only clear duplicate-like pair is the Qwen3.5 save_pretrained regression, where the later report narrows the same serialization problem to visual encoder keys.", + "confidence": 0.87, + "canonical_issue_reason": "issue:45216 is the broadest and earliest report of the Qwen3.5 save_pretrained regression, with a clear version boundary and the core serialization failure; issue:45357 reads as a later, narrower follow-up on the same bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45812 has the strongest mix of scope and specificity among the issues here, making it the cleanest anchor for this loose cluster.", + "best_issue_reason": "issue:45216 best captures the shared underlying bug because it describes the save_pretrained regression at the right abstraction level, while 45357 is a specific symptom of that same failure.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44849", "right": "issue:45325", "accept": false, - "reason": "Different model families and bugs: Qwen3.5 hidden-state behavior vs Qwen2.5-VL rope index scaling." + "reason": "Different Qwen model families and different bugs: hidden-states output logic vs RoPE position scaling for still images." }, { "left": "issue:44464", "right": "issue:45987", "accept": false, - "reason": "Both involve generation/cache behavior, but one is chunked generation with compiled forward and the other is a StaticCache API contract issue." + "reason": "Unrelated failures: chunked generation inconsistency with compiled forward vs StaticCache length return-type contract." }, { "left": "issue:43262", "right": "issue:43377", "accept": false, - "reason": "Audio chat-template sample-rate default vs MIMI encoder padding-mask correctness; different subsystems and symptoms." + "reason": "Both audio-related, but one is a chat-template sampling-rate default and the other is a batching/padding-mask correctness bug in MIMI." }, { "left": "issue:44987", "right": "issue:44991", "accept": false, - "reason": "Both are loading regressions, but they affect different artifacts and failure modes: a model load break vs tokenizer loading for a different checkpoint." + "reason": "Both are model-loading regressions, but they affect different models and appear to stem from different code paths." }, { "left": "issue:43646", "right": "issue:44534", "accept": false, - "reason": "Custom model initialization breakage is not the same bug as non-persistent buffers being filled with junk." + "reason": "Custom model initialization breakage in notebooks is a different issue from non-persistent buffers being filled with junk." }, { "left": "issue:43883", "right": "issue:44991", "accept": false, - "reason": "Molmo missing all_tied_weights_keys is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + "reason": "Molmo tied-weights attribute error is unrelated to tokenizer loading for EMBEDDIA/est-roberta." }, { "left": "issue:43404", "right": "issue:44466", "accept": false, - "reason": "Tie-weights / serialization issues are related in theme but not the same bug: one is missing lm_head tying, the other is device-dependent save serialization." + "reason": "Both mention tied weights/lm_head, but one is a model-specific missing tie and the other is a save_pretrained serialization inconsistency." }, { "left": "issue:44403", "right": "issue:45081", "accept": false, - "reason": "Unnecessary loading noise is a logging issue; _patch_mistral_regex crashing is a tokenizer patch failure." + "reason": "Generic loading noise is not the same as the Mistral regex patch crash." }, { "left": "issue:42915", "right": "issue:45084", "accept": false, - "reason": "Qwen3Moe FP8 config failure and non-template-node compile error are unrelated." + "reason": "Qwen3Moe FP8 config failure and non-template-node compile error are different concrete problems." }, { "left": "issue:43994", "right": "issue:44610", "accept": false, - "reason": "Different vision-model bugs: nonsensical SigLIP2 outputs vs OmDet-Turbo processor/model input-size mismatch." + "reason": "Different vision issues: nonsensical SigLIP2 outputs vs wrong processor image size for OmDet-Turbo." }, { "left": "issue:43377", "right": "issue:43819", "accept": false, - "reason": "Both are audio-related, but one is missing padding-mask support in MIMI and the other is DAC.from_latents diverging from forward due to missing STE." + "reason": "Different audio codec/model bugs: padding-mask batching mismatch vs DAC latent/STE mismatch." }, { "left": "issue:43322", "right": "issue:43540", "accept": false, - "reason": "Llava Next loading segfault and Qwen3OmniMoe video ValueError are different multimodal failure paths." + "reason": "Different multimodal models and failure modes: Llava Next loading segfault vs Qwen3OmniMoe video-processing ValueError." }, { "left": "issue:28530", "right": "issue:34567", "accept": false, - "reason": "Trainer early-stopping metric selection and num_input_tokens_seen not updating are separate TrainerState/training-loop issues." + "reason": "Trainer early-stopping configuration issue and token-count tracking bug are unrelated." }, { "left": "issue:43645", "right": "issue:44534", "accept": false, - "reason": "Notebook custom-model init failure is not the same as v5 junk-filled non-persistent buffers." + "reason": "Notebook custom-model initialization regression is separate from the non-persistent-buffer serialization bug." }, { "left": "issue:43089", "right": "issue:45987", "accept": false, - "reason": "Generation overhead/performance issue vs cache length return-type contract bug; different level and code path." + "reason": "Generation performance overhead and StaticCache return-type mismatch are not the same bug." }, { "left": "issue:44451", "right": "issue:45081", "accept": false, - "reason": "ScandiBERT loading failure and Mistral regex patch crash are unrelated model/tokenizer bugs." + "reason": "ScandiBERT loading failure is unrelated to the Mistral regex patch crash." }, { "left": "issue:44987", "right": "issue:45081", "accept": false, - "reason": "A model-loading regression and a tokenizer regex-patch crash are different failures." + "reason": "Different models and different failure points; no shared underlying bug is evident." }, { "left": "issue:44265", "right": "issue:44610", "accept": false, - "reason": "torch.export failure with torch_compilable_check is unrelated to OmDet-Turbo image-size preprocessing." + "reason": "torch.export/torch_compilable_check failure is unrelated to an image-size processor bug." }, { "left": "issue:44220", "right": "issue:44683", "accept": false, - "reason": "Audio feature extraction and compiled flex_attention on torch>=2.9 are separate issues." + "reason": "Feature-extraction helper bug and compiled flex_attention failure are unrelated." }, { "left": "issue:44991", "right": "issue:46315", "accept": false, - "reason": "Tokenizer load failure for a specific model vs slow-tokenizer performance regression in convert_tokens_to_ids; different bugs." + "reason": "Tokenizer loading failure and slow convert_tokens_to_ids performance are different issues." }, { "left": "issue:45216", "right": "issue:45357", "accept": true, - "reason": "Both describe the same Qwen3.5 save_pretrained regression: incorrect serialized checkpoint/visual encoder keys across versions." + "reason": "Same Qwen3.5 save_pretrained regression: 45357 is a narrower follow-up describing incorrect visual encoder keys, which fits the same underlying serialization bug." }, { "left": "issue:42915", "right": "issue:44291", "accept": false, - "reason": "Qwen3Moe FP8 config failure and init_empty_weights handling of _is_hf_initialized are unrelated initialization bugs." + "reason": "Different incompatibilities: Qwen3Moe FP8 config failure vs init_empty_weights receiving an unexpected argument." }, { "left": "issue:36032", "right": "issue:43742", "accept": false, - "reason": "T5 tokenizer method-name conflict and MobileLLM key error during load are distinct tokenizer/model-loading issues." + "reason": "Tokenizer method-name conflict and MobileLLM key error are unrelated." }, { "left": "issue:42915", "right": "issue:44589", "accept": false, - "reason": "Qwen3Moe FP8 config failure and missing Float8 storage lookup are different torch/quantization problems." + "reason": "Different FP8-related failures with different symptoms and likely different code paths." }, { "left": "issue:40444", "right": "issue:46326", "accept": false, - "reason": "Multi-image IterableDataset failure and mistral_common continue_final_message bug are different multimodal/tokenizer behaviors." + "reason": "IterableDataset multi-image finetuning failure is unrelated to mistral_common continue_final_message behavior." }, { "left": "issue:44062", "right": "issue:44987", "accept": false, - "reason": "AddedToken special-argument duplication is unrelated to loading physical-intelligence/fast." + "reason": "AddedToken constructor argument conflict and a model-loading regression for physical-intelligence/fast are unrelated." }, { "left": "issue:43531", "right": "issue:44877", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and strict granite_speech config loading are different issues." + "reason": "Qwen3-MoE sliding_window behavior and granite_speech strict-config loading are different problems." }, { "left": "issue:43611", "right": "issue:44534", "accept": false, - "reason": "base_model_prefix load breakage is not the same as junk-filled non-persistent buffers." + "reason": "base_model_prefix loading breakage is separate from junk-filled non-persistent buffers." }, { "left": "issue:45812", "right": "issue:46491", "accept": false, - "reason": "Both are tokenizer regressions, but Granite wrong token IDs and CodeLlama leading-space round-trip loss are different model-specific bugs." + "reason": "Both tokenizer regressions, but for different model families and different observable failures." }, { "left": "issue:43425", "right": "issue:44403", "accept": false, - "reason": "Torch 2.10 compatibility and noisy model loading are unrelated." + "reason": "Torch version compatibility and generic transformer-loading noise do not describe the same bug." }, { "left": "issue:43296", "right": "issue:43531", "accept": false, - "reason": "PaddleOCR-VL load failure in vLLM and Qwen3-MoE sliding_window issue are unrelated." + "reason": "PaddleOCR-VL/vLLM loading failure is unrelated to the Qwen3-MoE sliding_window issue." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is conservative and grounded in the packet titles. The single accepted pair (issue:45216 / issue:45357) looks like a plausible duplicate/follow-up on the same Qwen3.5 save_pretrained regression, and the other soft-edge rejections do not overclaim overlap." + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted soft pair, issue:45216 vs issue:45357, is plausibly the same Qwen3.5 save_pretrained regression, with 45357 reading as a later/narrower follow-up. The rest of the pair verdicts are appropriately cautious and the summary does not overclaim broader duplication beyond what the titles support." }, "evaluator_used": true, "retried": false, @@ -3630,8 +3552,8 @@ "issue:44291|issue:44589", "issue:44829|issue:45479", "issue:44079|issue:45663", - "issue:40444|issue:45325", "issue:43482|issue:45538", + "issue:40444|issue:45325", "issue:43482|issue:43525", "issue:41553|issue:43582", "issue:43317|issue:43856", @@ -3644,20 +3566,20 @@ "issue:41720|issue:45081", "issue:43531|issue:44291" ], - "prepared_review_unit_hash": "8c57eba92d168410af7410354b9c919370cd98f9622ab7b3a7cea0d46456a66e", + "prepared_review_unit_hash": "bc30854faca267c4db5aba5a3acc1b040e1e7aeaec6c33ab3c301132643c77a0", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 36296, + "serialized_chars": 36295, "estimated_input_tokens": 9074, "estimated_eval_tokens": 18404 }, @@ -3666,186 +3588,189 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is a mixed-topic cluster spanning unrelated tokenizer, loading, training, attention, and hardware/offload bugs. None of the suggested soft pairs look like the same concrete underlying issue, so there is no real duplicate canon here.", - "confidence": 0.93, - "canonical_issue_reason": null, + "summary": "No soft pair is strong enough to treat as a confirmed duplicate. The Qwen2.5-VL pair (44479/45325) is thematically related, but the packet marks it deterministic_accept=false and does not establish that 45325 is the root cause of 44479; it should be treated as a tentative topical relation only.", + "confidence": 0.83, + "canonical_issue_reason": "issue:44479 is the best umbrella issue for the Qwen video-regression topic because it describes the broader user-facing failure across multiple Qwen VL variants. However, its relationship to issue:45325 is only tentative in the packet, not a confirmed duplicate/root-cause pairing.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:44479 is the strongest triage anchor among the cluster because it covers the widest visible impact in the shared Qwen VL video-position-id regression theme, even though the evidence is not sufficient to collapse it with 45325.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43632", "right": "issue:44756", "accept": false, - "reason": "Different bugs: `_is_hf_initialized` regression vs Strix Halo mmap OOM mitigation." + "reason": "Different bugs: _is_hf_initialized / version regression vs mmap OOM on Strix Halo." }, { "left": "issue:36032", "right": "issue:41720", "accept": false, - "reason": "Tokenizer method-name conflict is unrelated to Qwen3 auto device-map CUDA asserts." + "reason": "Tokenizer attribute conflict and Qwen3 auto device-map CUDA assert are unrelated." }, { "left": "issue:44479", "right": "issue:45325", "accept": false, - "reason": "Both are Qwen2.5-VL/video-related, but one is a broad v5.3 video regression while the other is still-image position-id scaling." + "reason": "Related Qwen2.5-VL video-position-id symptoms, but the packet does not conservatively prove the same underlying bug; deterministic_accept is false." }, { "left": "issue:44977", "right": "issue:46326", "accept": false, - "reason": "Flash-attention generation failure is unrelated to mistral_common `continue_final_message` tokenizer behavior." + "reason": "Both are generation/tokenizer-adjacent, but the concrete failures and code paths differ." }, { "left": "issue:43792", "right": "issue:44589", "accept": false, - "reason": "Whisper runtime failure and missing Float8 storage are unrelated code paths." + "reason": "Whisper runtime failure and Float8 storage lookup are unrelated." }, { "left": "issue:44403", "right": "issue:44991", "accept": false, - "reason": "Generic loading noise vs a specific tokenizer load regression for EMBEDDIA/est-roberta." + "reason": "Generic loading noise vs a specific tokenizer-loading regression for a model repo." }, { "left": "issue:40990", "right": "issue:44811", "accept": false, - "reason": "GPT-OSS perplexity issue is unrelated to Whisper `batch_decode` skip_special_tokens handling." + "reason": "Perplexity on gpt-oss-20b is unrelated to Whisper batch_decode skip_special_tokens behavior." }, { "left": "issue:33666", "right": "issue:40444", "accept": false, - "reason": "Both involve Qwen2-VL training, but one is multi-GPU training and the other is iterable-dataset multi-image handling." + "reason": "Both are Qwen-VL related, but the training setups and failure modes are different." }, { "left": "issue:44683", "right": "issue:45020", "accept": false, - "reason": "Compiled flex_attention on torch 2.9 and remote_code breakage are distinct failures." + "reason": "Torch flex_attention compilation and remote_code loading regressions are separate problems." }, { "left": "issue:43381", "right": "issue:45923", "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to a missing `supports_gradient_checkpointing` flag." + "reason": "Eval-mode gradient checkpointing limitation is not the same as a missing supports_gradient_checkpointing flag." }, { "left": "issue:43632", "right": "issue:44534", "accept": false, - "reason": "Different regression classes: `_is_hf_initialized` flag handling vs non-persistent buffer junk values." + "reason": "Different internal regressions with different failure surfaces." }, { "left": "issue:43421", "right": "issue:44568", "accept": false, - "reason": "Runtime post-processor updates and missing BOS/EOS insertion are separate tokenizer bugs." + "reason": "Both involve special tokens, but one is runtime post-processor updates and the other is BOS/EOS insertion in a specific tokenizer." }, { "left": "issue:44291", "right": "issue:44589", "accept": false, - "reason": "`init_empty_weights` argument mismatch is unrelated to Float8 storage lookup failure." + "reason": "Unexpected _is_hf_initialized argument and missing Float8 storage object are unrelated loading errors." }, { "left": "issue:44829", "right": "issue:45479", "accept": false, - "reason": "Both are training pathologies, but flash-attention-3 degeneration and single-label zero-loss are different root causes." + "reason": "Both mention degenerate training, but the root causes differ (flash_attention_3 vs classification-loss setup)." }, { "left": "issue:44079", "right": "issue:45663", "accept": false, - "reason": "ModelOutput key assignment and Gemma4 FSDP2 shared_kv_states are unrelated." + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states rebuilding are unrelated." }, { - "left": "issue:40444", - "right": "issue:45325", + "left": "issue:43482", + "right": "issue:45538", "accept": false, - "reason": "Multi-image iterable dataset failure is not the same as still-image temporal position scaling." + "reason": "GGUF loading failure and CLIPTokenizer model_max_length are separate tokenizer/model-loading issues." }, { - "left": "issue:43482", - "right": "issue:45538", + "left": "issue:40444", + "right": "issue:45325", "accept": false, - "reason": "Qwen2.5-GGUF v5 loading failure is unrelated to CLIPTokenizer max-length behavior." + "reason": "Both are Qwen2.5-VL-related, but one is multi-image finetuning and the other is temporal position-id scaling." }, { "left": "issue:43482", "right": "issue:43525", "accept": false, - "reason": "GGUF loading failure and missing `pad_token_id` on Llama4Config are distinct issues." + "reason": "GGUF loading failure is unrelated to Llama4Config missing pad_token_id." }, { "left": "issue:41553", "right": "issue:43582", "accept": false, - "reason": "Bad Voxtral error messaging is unrelated to Apple Silicon `caching_allocator_warmup` TypeError." + "reason": "AutoTokenizer error messaging for Voxtral is unrelated to Apple Silicon allocator warmup TypeError." }, { "left": "issue:43317", "right": "issue:43856", "accept": false, - "reason": "GPU+CPU offload failure for dequantized models is unrelated to Qwen3 MoE training memory inefficiency." + "reason": "GPU/CPU offload loading failure and Qwen3 MoE memory usage are different bugs." }, { "left": "issue:42175", "right": "issue:45542", "accept": false, - "reason": "Package-extras backend omission and undefined tf backend from tensorboard-only install are different packaging bugs." + "reason": "Both touch TensorFlow packaging, but they describe different install states and failures." }, { "left": "issue:43818", "right": "issue:46402", "accept": false, - "reason": "Video-LLaVA temporal-attention/weight-sharing bug is unrelated to Sam3Model text encoder weight loading." + "reason": "Video-LLaVA tower architecture issue and Sam3Model text-encoder loading are unrelated." }, { "left": "issue:42796", "right": "issue:43746", "accept": false, - "reason": "BPE marker handling in a tokenizer is unrelated to GraniteSpeech PEFT checkpoint loading." + "reason": "Tokenizer BPE-marker handling and GraniteSpeech PEFT checkpoint loading are unrelated." }, { "left": "issue:44462", "right": "issue:45812", "accept": false, - "reason": "Ignoring `tokenizer.json` and Granite token-ID regression are both tokenizer issues but not the same concrete bug." + "reason": "Both involve AutoTokenizer, but the repo-tokenizer precedence bug and Granite token-ID regression are not the same code-path." }, { "left": "issue:43818", "right": "issue:45357", "accept": false, - "reason": "Video-LLaVA architecture bug is unrelated to incorrect visual encoder keys in Qwen3.5 save_pretrained." + "reason": "Video-LLaVA tower issue and Qwen3.5 save_pretrained visual-key regression are different regressions." }, { "left": "issue:43377", "right": "issue:45491", "accept": false, - "reason": "Batched-vs-single MIMI encoder mismatch from padding masks is different from Gemma3 NaNs with sliding-window all-padding windows." + "reason": "MIMI padding-mask batching and Gemma3 sliding-window NaNs are different model bugs." }, { "left": "issue:41720", "right": "issue:45081", "accept": false, - "reason": "Qwen3 device-map CUDA assert and Mistral regex patch AttributeError do not share the same code path." + "reason": "CUDA assert during auto device mapping is unrelated to the Mistral tokenizer regex crash." }, { "left": "issue:43531", "right": "issue:44291", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and `_is_hf_initialized` init_empty_weights handling are unrelated." + "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights/_is_hf_initialized loading errors are separate issues." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The summary overstates the evidence. In the packet, the only Qwen pair (issue:44479 / issue:45325) is explicitly marked deterministic_accept=false, so treating it as a confirmed duplicate/canonical pair is not conservative. The titles do show related Qwen2.5-VL/video-position-id behavior, but the claim that 45325 is the root cause behind 44479 is not grounded by the packet. The other soft-edge rejections are generally conservative enough, but the accepted pair and the canonical-issue rationale should be downgraded to a tentative thematic relation at most." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -3946,20 +3871,20 @@ "issue:42796|issue:44977", "issue:43377|issue:44610" ], - "prepared_review_unit_hash": "38206bf5973bb312bc968629e238dac151ac9d5e88a6b7e8f7f1a2e837aaac00", + "prepared_review_unit_hash": "f2f711e03ef1105d6f17c4e15667008c8acb30a7b9fac4dab547f8e0b0256db5", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 37463, + "serialized_chars": 37461, "estimated_input_tokens": 9366, "estimated_eval_tokens": 18988 }, @@ -3968,204 +3893,204 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is mostly a grab-bag of unrelated regression reports; I found no convincing duplicate pairs. The strongest representative issue is the broad tokenizer regression in issue 45920.", - "confidence": 0.79, - "canonical_issue_reason": "Issue 45920 is the broadest open regression here, covering wrong AutoTokenizer IDs across multiple model families and serving as the clearest umbrella bug in the set.", + "summary": "The set is mostly heterogeneous, model-specific regressions. No soft pair looks like a clear duplicate; the closest ones are related by theme but not the same concrete bug. The best global issue anchor is the broad AutoTokenizer regression (issue 45920), since it is active, has inbound references, and covers a wide cross-model tokenizer failure mode.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45920 is the broadest and most central open regression here: it affects many models, has inbound references, and describes a reusable tokenizer failure pattern rather than a single model quirk.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45920 is the best representative issue because it is open, broadly scoped, and already aggregates multiple affected model families.", + "best_issue_reason": "Issue 45920 is the best cluster anchor because it is the most general, highest-impact issue in the set and is already being referenced externally.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43611", "right": "issue:43646", "accept": false, - "reason": "Both are v5 model-loading regressions, but one is about base_model_prefix loading and the other about custom initialization; the concrete code paths differ." + "reason": "Both are v5 loading regressions, but one is about `base_model_prefix` and the other about custom model initialization; different failure points and likely different fixes." }, { "left": "issue:44265", "right": "issue:45072", "accept": false, - "reason": "torch.export/export-check failure and bfloat16 dtype mismatch are different bugs affecting different subsystems." + "reason": "Unrelated bugs: `torch.export`/`torch_compilable_check` vs a bfloat16 dtype mismatch in inference." }, { "left": "issue:42175", "right": "issue:43296", "accept": false, - "reason": "Package backend installation is unrelated to a vLLM model-loading failure." + "reason": "Different subsystems and symptoms: packaging/backend selection vs loading a specific model in vLLM." }, { "left": "issue:43994", "right": "issue:44898", "accept": false, - "reason": "These involve different models and different symptoms: SigLIP2 docs/output issues vs Perceiver resolution/interpolation failure." + "reason": "Both are vision-model regressions, but they target different models and different behaviors (nonsensical outputs vs image-size handling)." }, { "left": "issue:43792", "right": "issue:43957", "accept": false, - "reason": "Whisper runtime failure and meta-device loading regressions are distinct issues." + "reason": "One is a model-specific runtime failure, the other is a meta-device loading regression; not the same code-path bug." }, { "left": "issue:45325", "right": "issue:45834", "accept": false, - "reason": "Qwen2.5-VL rope-index scaling and Kosmos2.5 OCR index errors are different model-specific bugs." + "reason": "Both involve multimodal position/index handling, but one is a Qwen2.5-VL rope-index scaling bug and the other is a Kosmos2.5 OCR length/index error." }, { "left": "issue:43454", "right": "issue:44466", "accept": false, - "reason": "One is about failing to tie lm_head weights; the other is about inconsistent lm_head.weight serialization. Related area, but not the same bug." + "reason": "Both mention `lm_head`, but one is about missing weight tying and the other about inconsistent serialization; related area, not the same bug." }, { "left": "issue:42915", "right": "issue:43482", "accept": false, - "reason": "Qwen3Moe FP8 config failure and GGUF loading failure are different loader/config paths." + "reason": "Different models and different loading/quantization paths; no shared concrete defect." }, { "left": "issue:44079", "right": "issue:44534", "accept": false, - "reason": "ModelOutput key assignment and non-persistent buffer initialization are unrelated regressions." + "reason": "`ModelOutput` key assignment and non-persistent buffer initialization are unrelated." }, { "left": "issue:43632", "right": "issue:45072", "accept": false, - "reason": "_is_hf_initialized breakage and dtype mismatch in inference are different failures." + "reason": "Different regressions: `_is_hf_initialized` handling vs dtype mismatches in inference." }, { "left": "issue:45325", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen2.5-VL position-id issues, but one is still-image scaling and the other is video-input position ids; not the same concrete bug." + "reason": "Both are Qwen2.5-VL position-id issues, but one is still-image temporal scaling and the other is video `vision_position_ids`; close but not clearly the same bug." }, { "left": "issue:40444", "right": "issue:45381", "accept": false, - "reason": "Multi-image finetuning failure and video position-id regression are different multimodal bugs." + "reason": "Both concern Qwen2.5-VL multimodal inputs, but one is finetuning with multiple images per prompt and the other is a video position-id regression." }, { "left": "issue:43644", "right": "issue:44079", "accept": false, - "reason": "Non-persistent buffer junk and ModelOutput key assignment are not the same code-path problem." + "reason": "One is buffer initialization junk, the other is `ModelOutput` field assignment; no duplicate-level overlap." }, { "left": "issue:44291", "right": "issue:44877", "accept": false, - "reason": "init_empty_weights/_is_hf_initialized TypeError and strict config loading are unrelated." + "reason": "Both are loading-related, but one is an `init_empty_weights` argument regression and the other is strict config validation for GraniteSpeech." }, { "left": "issue:44479", "right": "issue:45823", "accept": false, - "reason": "Qwen video-input regression and Gemma4 device_map mismatch are different model-specific forward failures." + "reason": "Both involve device/runtime issues, but they affect different models and different mechanisms (`device_map`/video regression vs Gemma4 PLE forward mismatch)." }, { "left": "issue:35707", "right": "issue:40444", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen2.5-VL multi-image dataset finetuning." + "reason": "Generation with `inputs_embeds`/`past_key_values` is unrelated to Qwen2.5-VL multi-image finetuning." }, { "left": "issue:43761", "right": "issue:44079", "accept": false, - "reason": "CLIPVisionModel hidden_states regression is not the same concrete bug as generic ModelOutput key assignment." + "reason": "Different failures: CLIP hidden states not returned vs `ModelOutput` key bookkeeping." }, { "left": "issue:42796", "right": "issue:45381", "accept": false, - "reason": "Tokenizer BPE-marker handling and Qwen2.5-VL video position ids are unrelated." + "reason": "Tokenizer BPE-marker handling is unrelated to Qwen2.5-VL video position IDs." }, { "left": "issue:43296", "right": "issue:45800", "accept": false, - "reason": "A PaddleOCR-VL load failure and a torch/transformers version incompatibility are different issues." + "reason": "A model-loading failure in vLLM is not the same as a broad torch/transformers version incompatibility report." }, { "left": "issue:43606", "right": "issue:44265", "accept": false, - "reason": "CPU offload device mismatch and torch.export failures are unrelated." + "reason": "CPU offload device mismatch and `torch.export` failure from `torch_compilable_check` are different bugs." }, { "left": "issue:43701", "right": "issue:45325", "accept": false, - "reason": "resume_from_checkpoint key mismatch has nothing to do with Qwen2.5-VL rope-index handling." + "reason": "Checkpoint key mismatch and Qwen2.5-VL rope-index scaling are unrelated." }, { "left": "issue:43881", "right": "issue:44488", "accept": false, - "reason": "These are different model-loading failures with no shared concrete code path." + "reason": "Different model-loading problems with no shared code-path evidence." }, { "left": "issue:43606", "right": "issue:43819", "accept": false, - "reason": "CPU offload device mismatch and DAC.from_latents/STE mismatch are distinct bugs." + "reason": "Device mismatch on CPU offload is unrelated to DAC latent reconstruction mismatch." }, { "left": "issue:43454", "right": "issue:43550", "accept": false, - "reason": "lm_head weight tying/serialization and SDPA torch.compile failure are unrelated." + "reason": "Weight-tying/generation correctness and torch.compile SDPA issues are different defects." }, { "left": "issue:39692", "right": "issue:40444", "accept": false, - "reason": "SigLIP2 documentation/model mismatch and Qwen2.5-VL multi-image finetuning failure are unrelated." + "reason": "Documentation/example errors are not the same bug as a finetuning failure with multiple images per prompt." }, { "left": "issue:41720", "right": "issue:45823", "accept": false, - "reason": "Both mention device mapping, but one is a CUDA assert on Qwen3 and the other is a Gemma4 forward device mismatch; not the same bug." + "reason": "Both mention device mapping, but the affected models and concrete failures differ substantially." }, { "left": "issue:44186", "right": "issue:44265", "accept": false, - "reason": "Tokenizer crashes on NER/padding are unrelated to torch.export and torch_compilable_check." + "reason": "Tokenizer padding/truncation crashes and `torch.export`/compilability issues are unrelated." }, { "left": "issue:46305", "right": "issue:46315", "accept": false, - "reason": "Streaming mask off-by-one and tokenizer conversion slowness are different problems." + "reason": "Sliding-window masking off-by-one and tokenizer slowdown are entirely different regressions." }, { "left": "issue:43262", "right": "issue:43746", "accept": false, - "reason": "Audio chat-template sample-rate handling and GraniteSpeech PEFT checkpoint loading are unrelated." + "reason": "Audio chat-template sample-rate defaults and GraniteSpeech PEFT checkpoint loading are separate issues." }, { "left": "issue:44568", "right": "issue:45920", "accept": false, - "reason": "Both are tokenizer regressions, but one is a special-token/BOS-EOS issue for a specific tokenizer while the other is a broader wrong-token-IDs regression across many models." + "reason": "Both are tokenizer-related v5 regressions, but one is special-token insertion for a specific tokenizer and the other is widespread wrong token IDs across many models." }, { "left": "issue:42796", "right": "issue:44977", "accept": false, - "reason": "Ministral tokenizer BPE-marker handling and Qwen3.5 flash-attention generation are unrelated." + "reason": "Tokenizer BPE-marker handling and Qwen3.5 flash-attention generation are unrelated." }, { "left": "issue:43377", "right": "issue:44610", "accept": false, - "reason": "MIMI padding-mask output mismatch and OmDet-Turbo processor/model resolution mismatch are different bugs." + "reason": "Padding-mask behavior in MIMI encoder and OmDet-Turbo processor size mismatch are different model pipeline bugs." } ] }, @@ -4196,7 +4121,6 @@ "issue:36331", "issue:40990", "issue:41720", - "issue:42175", "issue:42617", "issue:42796", "issue:42915", @@ -4220,8 +4144,10 @@ "issue:43883", "issue:44008", "issue:44186", + "issue:44387", "issue:44423", "issue:44466", + "issue:44756", "issue:44877", "issue:44898", "issue:44977", @@ -4231,10 +4157,10 @@ "issue:45406", "issue:45584", "issue:45701", + "issue:45704", "issue:45759", "issue:45823", "issue:45923", - "issue:46291", "issue:46402" ], "soft_pairs": [ @@ -4248,7 +4174,6 @@ "issue:43873|issue:45701", "issue:43454|issue:43720", "issue:42796|issue:43854", - "issue:44991|issue:46291", "issue:43526|issue:43883", "issue:44466|issue:45357", "issue:41720|issue:44977", @@ -4266,224 +4191,225 @@ "issue:43550|issue:43720", "issue:43577|issue:43708", "issue:44008|issue:45823", - "issue:42175|issue:46291", "issue:36331|issue:42617", - "issue:44423|issue:45406" + "issue:44423|issue:45406", + "issue:44387|issue:45704", + "issue:43873|issue:44756" ], - "prepared_review_unit_hash": "56377588e9935bbad169d0f069e06e43273bde38bb6d60963f9e4e841d838e47", + "prepared_review_unit_hash": "6303265401f4824c1b9d3a909b88bc408bb8ab171260534512ce78630743781a", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, + "node_count": 48, + "item_count": 48, "soft_pair_count": 31, - "serialized_chars": 36651, - "estimated_input_tokens": 9163, - "estimated_eval_tokens": 18582 + "serialized_chars": 37190, + "estimated_input_tokens": 9298, + "estimated_eval_tokens": 18852 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a noisy mix of unrelated bug reports across trainer behavior, tokenizers, model loading, quantization/offloading, and multimodal processors. None of the soft pairs look like true duplicates; they mostly share broad subsystem themes rather than the same underlying bug.", - "confidence": 0.79, - "canonical_issue_reason": "If a representative issue is needed, 43873 is the broadest umbrella report in the offloading/quantization area and has inbound references, so it best stands in for that subtheme.", + "summary": "This cluster is highly heterogeneous and does not look like a duplicate set; the only recurring theme is a loose family of loading/offloading/quantization-related reports, with no pair meeting a strict same-bug threshold.", + "confidence": 0.33, + "canonical_issue_reason": "issue:43873 is the most central of the bunch because it is the broadest loading/offloading report and it appears in multiple soft-similarity links, but the cluster overall is too mixed to treat it as a true duplicate anchor.", "canonical_pr_reason": null, - "best_issue_reason": "43873 is the most central and general issue here: it describes the broader offloading-with-quantization failure, which is the closest thing to a cluster anchor among otherwise unrelated reports.", + "best_issue_reason": "issue:43873 is the best single issue to represent the cluster\u2019s loose offloading/quantization theme, but it is not a good global duplicate canonical because most other issues are unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:40990", "right": "issue:45584", "accept": false, - "reason": "Different models and failure modes: perplexity regression vs Whisper empty-transcription handling." + "reason": "Both are model behavior bugs, but one is perplexity on gpt-oss-20b and the other is Whisper empty-transcription handling; no shared code-path is evident." }, { "left": "issue:43454", "right": "issue:45072", "accept": false, - "reason": "One is tied-weights serialization, the other is a bfloat16 dtype mismatch in inference." + "reason": "Different failures: tied lm_head weights in AyaVision vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:43317", "right": "issue:43873", "accept": false, - "reason": "Related broad area, but one is a specific dequantized-model load/offload failure and the other is a generic quantized offloading report; not the same code-path bug." + "reason": "Both mention offloading/quantization, but one is dequantized model loading with device_map=auto and the other is a broader CUDA memory/offloading complaint; too generic to merge." }, { "left": "issue:43257", "right": "issue:45923", "accept": false, - "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is unrelated to a missing gradient-checkpointing flag." + "reason": "Accelerate+DeepSpeed MOE weight conversion and a missing gradient-checkpointing flag are unrelated issues." }, { "left": "issue:43377", "right": "issue:44898", "accept": false, - "reason": "MIMI padding-mask batching bug vs Perceiver interpolation issue; different components and symptoms." + "reason": "Different processors and different bugs: padding-mask-dependent encoder outputs vs image classification with interpolate_pos_encoding." }, { "left": "issue:40990", "right": "issue:42947", "accept": false, - "reason": "Perplexity regression and gradient-checkpointing behavior are unrelated." + "reason": "Perplexity/eval-quality issue vs gradient-checkpointing ineffectiveness with PEFT LoRA; not the same bug." }, { "left": "issue:43572", "right": "issue:45759", "accept": false, - "reason": "Missing config field in StableLm vs composite multimodal config unwrapping; different bugs." + "reason": "Both concern config/model loading, but one is a missing pad token field and the other is unwrapping composite text_config; different code paths." }, { "left": "issue:43873", "right": "issue:45701", "accept": false, - "reason": "Offloading/quantization issue vs version-dependent tokenization change; no shared concrete failure path." + "reason": "Offloading/quantization problems are not the same as version-dependent tokenization changes." }, { "left": "issue:43454", "right": "issue:43720", "accept": false, - "reason": "Tied-weight serialization bug vs packed-weight unpacking during accelerate loading; distinct mechanisms." + "reason": "AyaVision tied-weight serialization and BitNet packed-weight unpacking are distinct model-loading bugs." }, { "left": "issue:42796", "right": "issue:43854", "accept": false, - "reason": "Tokenizer BPE marker handling vs GLM-4.7 model loading failure; not the same bug." - }, - { - "left": "issue:44991", - "right": "issue:46291", - "accept": false, - "reason": "Tokenizer loading regression and kernels import/version requirement are unrelated." + "reason": "Tokenizer BPE-marker handling is unrelated to GLM-4.7 model loading failures." }, { "left": "issue:43526", "right": "issue:43883", "accept": false, - "reason": "ImageProcessor label reduction bug vs missing tied-weights attribute on a model class." + "reason": "BEiT label reduction and missing all_tied_weights_keys are unrelated implementation bugs." }, { "left": "issue:44466", "right": "issue:45357", "accept": false, - "reason": "Different Qwen3.5 serialization problems affecting different keys and behaviors." + "reason": "Both involve save/load behavior, but one is tied lm_head serialization and the other is incorrect visual encoder keys; different model components." }, { "left": "issue:41720", "right": "issue:44977", "accept": false, - "reason": "CUDA assert during auto device mapping vs flash-attention generation bug; different code paths." + "reason": "Both are Qwen-family generation issues, but one is an auto-device-map CUDA assert and the other is a flash-attention generation bug; not clearly the same defect." }, { "left": "issue:43881", "right": "issue:44991", "accept": false, - "reason": "GLM-4V model load failure vs tokenizer load regression for a different model." + "reason": "Different failure modes in model/tokenizer loading for unrelated models." }, { "left": "issue:26413", "right": "issue:28530", "accept": false, - "reason": "Checkpoint resume iterator failure vs early-stopping metric lookup; both trainer-related but not the same defect." + "reason": "Checkpoint resume/epoch iterator failure is unrelated to early stopping metric selection." }, { "left": "issue:42617", "right": "issue:42915", "accept": false, - "reason": "3D parallel script execution issue vs Qwen3 MoE FP8 config failure; unrelated." + "reason": "3D parallel execution failure and Qwen3Moe fine-grained FP8 loading failure are different problems." }, { "left": "issue:35707", "right": "issue:41720", "accept": false, - "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 device-mapping CUDA asserts." + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 auto device-map CUDA assertion." }, { "left": "issue:43065", "right": "issue:46402", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder vs Sam3 text encoder weight loading; same model family but different bugs." + "reason": "A dummy Conv2d in Sam3PixelDecoder and missing Sam3Model text-encoder weights are different Sam3 bugs." }, { "left": "issue:34634", "right": "issue:35707", "accept": false, - "reason": "BarkProcessor voice_preset issue and progressive generation bug do not share the same underlying code path." + "reason": "BarkProcessor voice_preset handling has no clear overlap with progressive generation using inputs_embeds." }, { "left": "issue:43450", "right": "issue:45701", "accept": false, - "reason": "Video processor batching shape bug vs tokenization version-change report; different subsystems." + "reason": "Video processor batch-shape bug and tokenization version-change bug are unrelated subsystems." }, { "left": "issue:44877", "right": "issue:44991", "accept": false, - "reason": "Strict config loading failure vs tokenizer loading regression; not the same issue." + "reason": "Strict config rejection for granite_speech is not the same as tokenizer loading failure for EMBEDDIA/est-roberta." }, { "left": "issue:43404", "right": "issue:45357", "accept": false, - "reason": "lm_head tying bug in Mistral3 vs incorrect visual encoder keys in Qwen3.5 save_pretrained; different serialization problems." + "reason": "Different tied-weight/save-pretrained issues affecting different model architectures." }, { "left": "issue:44186", "right": "issue:45072", "accept": false, - "reason": "Tokenizer padding/truncation crash vs dtype mismatch in bfloat16 inference; unrelated." + "reason": "Tokenizer crash on NER/batched padding is unrelated to dtype mismatch in inference." }, { "left": "issue:36331", "right": "issue:42915", "accept": false, - "reason": "CustomTrainer compute_loss signature change vs Qwen3 MoE weight conversion failure; unrelated." + "reason": "CustomTrainer API breakage and Qwen3Moe FP8 loading failure are unrelated." }, { "left": "issue:43550", "right": "issue:43720", "accept": false, - "reason": "torch.compile/SDPA failure in Bamba vs accelerate loading/unpacking issue in BitNet." + "reason": "torch.compile+SDPA failure in Bamba is not the same as accelerate loading not unpacking BitNet weights." }, { "left": "issue:43577", "right": "issue:43708", "accept": false, - "reason": "Model dtype propagation bug vs resume_from_checkpoint max_steps miscalculation; different behaviors." + "reason": "Blip2 dtype not propagating and resume_from_checkpoint max_steps miscalculation are unrelated bugs." }, { "left": "issue:44008", "right": "issue:45823", "accept": false, - "reason": "Gemma3n variable-name collision in forward vs Gemma4 device mismatch with auto mapping; different models and defects." - }, - { - "left": "issue:42175", - "right": "issue:46291", - "accept": false, - "reason": "Backend packaging/install issue vs kernels dependency version/revision requirement; unrelated." + "reason": "Gemma 3n variable-name collision in forward is unrelated to Gemma4 PLE device placement mismatch." }, { "left": "issue:36331", "right": "issue:42617", "accept": false, - "reason": "Trainer compute_loss API break vs inability to run a parallel script; no shared bug." + "reason": "Trainer compute_loss API breakage and 3D parallel execution failure do not share a concrete code path." }, { "left": "issue:44423", "right": "issue:45406", "accept": false, - "reason": "Serve continuous-batching multimodal crash vs Gemma4Processor missing _tokenizer; different failure points." + "reason": "Both affect serve, but the first is multimodal continuous-batching with a string/.to error and the second is a missing _tokenizer on Gemma4Processor; different root causes." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Int4 CUDA reserved-memory growth and an apex RMSNorm memory leak are different memory issues with different causes." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Quantization/offloading behavior is too broad a commonality; disabling mmap on Strix Halo is a platform-specific OOM workaround, not the same bug." } ] }, @@ -4522,8 +4448,7 @@ "issue:43316", "issue:43317", "issue:43322", - "issue:43425", - "issue:43525", + "issue:43421", "issue:43526", "issue:43531", "issue:43540", @@ -4534,30 +4459,30 @@ "issue:43720", "issue:43749", "issue:43782", - "issue:43873", "issue:43881", "issue:44186", "issue:44361", - "issue:44387", - "issue:44530", + "issue:44492", + "issue:44509", "issue:44534", "issue:44610", "issue:44671", "issue:44743", - "issue:44756", "issue:44898", "issue:44913", "issue:45161", + "issue:45357", + "issue:45406", + "issue:45464", "issue:45632", "issue:45684", "issue:45701", - "issue:45704", + "issue:45759", "issue:46050", - "issue:46291" + "issue:46291", + "issue:46402" ], "soft_pairs": [ - "issue:44387|issue:45704", - "issue:43873|issue:44756", "issue:43749|issue:44534", "issue:43550|issue:43606", "issue:44610|issue:44898", @@ -4569,11 +4494,10 @@ "issue:41720|issue:43322", "issue:41720|issue:42617", "issue:40990|issue:45701", - "issue:43425|issue:46291", "issue:36010|issue:42617", "issue:41720|issue:45161", - "issue:43525|issue:46291", "issue:43531|issue:43572", + "issue:42915|issue:46291", "issue:28530|issue:33666", "issue:43526|issue:43881", "issue:44671|issue:44743", @@ -4585,216 +4509,252 @@ "issue:43550|issue:43653", "issue:43606|issue:43720", "issue:30064|issue:34634", - "issue:44530|issue:46291" + "issue:45406|issue:45464", + "issue:43421|issue:45701", + "issue:42915|issue:43322", + "issue:45357|issue:46402", + "issue:44492|issue:44509", + "issue:43606|issue:43653", + "issue:43531|issue:45759" ], - "prepared_review_unit_hash": "f64305b15a8ea1780fbdf52655989cc70d87c9e5a0121a6e4a97be8f287d48f6", + "prepared_review_unit_hash": "dc0c2b7e41e8670db7e921d9d2edf9f4ddfe8cff7da5857c7eebfeed8719bc47", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 30, - "serialized_chars": 35720, - "estimated_input_tokens": 8930, - "estimated_eval_tokens": 18116 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37345, + "estimated_input_tokens": 9337, + "estimated_eval_tokens": 18930 }, "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 (tokenizers, model configs, offloading, vision processors, and a few loader regressions). The soft links are mostly topical similarity rather than true duplicates, so I would reject them all.", - "confidence": 0.58, - "canonical_issue_reason": "Issue 43749 is the strongest anchor: it has high discussion/references and is a concrete loading-related regression, making it the best representative among these otherwise heterogeneous reports.", + "summary": "This cluster is mostly semantic noise: the issues span unrelated models, tokenizer behaviors, loading paths, docs, and runtime errors. None of the soft pairs look like the same underlying bug/change, so no duplicate consolidation is warranted.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 43749 is the most substantial and well-documented bug report in the set, with the clearest regression-style description and the most triage activity.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44387", - "right": "issue:45704", - "accept": false, - "reason": "Both mention memory/quantization side effects, but one is int4/OOM and the other is an Apex RMSNorm leak in T5; different code paths and fixes." - }, - { - "left": "issue:43873", - "right": "issue:44756", - "accept": false, - "reason": "Both involve memory/offload, but one is quantization offloading behavior and the other is mmap on Strix Halo causing OOM; not the same bug." - }, { "left": "issue:43749", "right": "issue:44534", "accept": false, - "reason": "One is broken FSDP CPU RAM efficient loading, the other is non-persistent buffers being filled with junk; related area, different failure." + "reason": "One is FSDP CPU/RAM-efficient loading breaking; the other is v5 non-persistent buffer initialization. Different failure modes and code paths." }, { "left": "issue:43550", "right": "issue:43606", "accept": false, - "reason": "torch.compile+SDPA on Bamba is unrelated to Bark CPU-offload device mismatch; only broad model-loading/runtime similarity." + "reason": "Different models and bugs: torch.compile+SDPA on Bamba vs CPU offload device mismatch on Bark." }, { "left": "issue:44610", "right": "issue:44898", "accept": false, - "reason": "Both are image-resolution related, but one is a processor/model input-size mismatch and the other is Perceiver interpolation failing; not the same bug." + "reason": "Both involve image resolution, but they are separate model-specific preprocessing bugs with different expected sizes and failure points." }, { "left": "issue:44186", "right": "issue:44610", "accept": false, - "reason": "Tokenizer NER/padding crash vs vision processor output-size mismatch; different components and symptoms." + "reason": "LayoutLMv2 tokenizer NER/padding crash is unrelated to OmDet-Turbo image-size mismatch." }, { "left": "issue:43295", "right": "issue:45701", "accept": false, - "reason": "Both involve tokenization behavior changes, but one is a regression around processor.tokenizer/image passing and the other is a broader version-dependent tokenization change." + "reason": "The first is a regression in processor/tokenizer access; the second is a broader tokenization-change report. Not the same concrete bug." }, { "left": "issue:45632", "right": "issue:45684", "accept": false, - "reason": "Cache-path collision for trust_remote_code is unrelated to read-only permission propagation in save_pretrained/register_for_auto_class." + "reason": "Cache-path collisions for trust_remote_code and read-only permission propagation in save_pretrained are unrelated." }, { "left": "issue:42915", "right": "issue:43317", "accept": false, - "reason": "FineGrainedFP8Config failure is not the same as dequantized model loading with gpu+cpu offload." + "reason": "Both are loading/quantization-related, but they describe different models, configs, and distinct offload failures." }, { "left": "issue:43322", "right": "issue:46050", "accept": false, - "reason": "Segfault loading Llava Next and sinq save/load corruption are different model families and failure modes." + "reason": "Segfault when loading Llava Next is not the same as sinq quantization save/load breakage." }, { "left": "issue:41720", "right": "issue:43322", "accept": false, - "reason": "Qwen3 auto device-map cuda assert is unrelated to Llava Next segmentation faults." + "reason": "Different model families and symptoms: cudaErrorAssert on Qwen3 vs segfault on Llava Next." }, { "left": "issue:41720", "right": "issue:42617", "accept": false, - "reason": "Qwen3 device-map assert and failure to run 3d_parallel.py are distinct issues." + "reason": "Qwen3 auto device mapping failure is unrelated to not being able to run 3d_parallel.py." }, { "left": "issue:40990", "right": "issue:45701", "accept": false, - "reason": "High perplexity on gpt-oss and tokenization version changes are only loosely related; no clear same bug." - }, - { - "left": "issue:43425", - "right": "issue:46291", - "accept": false, - "reason": "Torch version incompatibility is a dependency/runtime issue, while kernels import requires version or revision is a package metadata/API issue." + "reason": "High perplexity on a specific model/dataset is not the same as a general tokenization-version change report." }, { "left": "issue:36010", "right": "issue:42617", "accept": false, - "reason": "GenerationMixin import error and 3d_parallel.py failure are unrelated." + "reason": "GenerationMixin import failure and a 3d_parallel.py runtime issue are unrelated." }, { "left": "issue:41720", "right": "issue:45161", "accept": false, - "reason": "Both are Qwen/GPT model issues, but one is device-map CUDA assert and the other is tensor-parallel support; different problems." + "reason": "Different Qwen/GPT-OSS-related loading issues with different symptoms and root causes." }, { - "left": "issue:43525", - "right": "issue:46291", + "left": "issue:43531", + "right": "issue:43572", "accept": false, - "reason": "Missing pad_token_id in Llama4Config is unrelated to kernels import requirements." + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to StableLmConfig missing pad_token_idx." }, { - "left": "issue:43531", - "right": "issue:43572", + "left": "issue:42915", + "right": "issue:46291", "accept": false, - "reason": "Qwen3-MoE sliding-window behavior and StableLmConfig pad_token_idx mismatch are different config/model bugs." + "reason": "FineGrainedFP8Config failure and pipelines import dependency-version failure are different problems." }, { "left": "issue:28530", "right": "issue:33666", "accept": false, - "reason": "Early-stopping metric selection and Qwen2-VL multi-GPU training are unrelated." + "reason": "Early stopping metric lookup and Qwen2-VL multi-GPU training are unrelated." }, { "left": "issue:43526", "right": "issue:43881", "accept": false, - "reason": "BEiT label reduction bug is unrelated to glm-4v loading failure." + "reason": "BEiT image label reduction bug is unrelated to glm-4v-9b loading failure." }, { "left": "issue:44671", "right": "issue:44743", "accept": false, - "reason": "CamemBERT masked-LM prediction regression and Qwen3 recurrent-state reset are different model internals." + "reason": "CamemBERT masked-LM prediction regression and Qwen3 modular recurrent-state reset are different code paths." }, { "left": "issue:44186", "right": "issue:44361", "accept": false, - "reason": "Two tokenizer bugs, but one is LayoutLMv2 NER/padding crash and the other is MLuke task AttributeError; not the same issue." + "reason": "Tokenizer crashes on NER/padding vs MLukeTokenizer task AttributeError: related subsystem, but not the same bug." }, { "left": "issue:43540", "right": "issue:43782", "accept": false, - "reason": "Video-input processing error in Qwen3OmniMoe differs from a from_pretrained weight_only loading error in Qwen3VL." + "reason": "Video-input processing error in Qwen3OmniMoe is different from weight_only load failure in Qwen3VL." }, { "left": "issue:43316", "right": "issue:44913", "accept": false, - "reason": "Gemma3TextConfig API discrepancy is not the same as GPTNeoX rotary_pct not persisting on reload." + "reason": "Gemma3TextConfig API discrepancy and GPTNeoX rotary_pct reload issue are unrelated config bugs." }, { "left": "issue:28530", "right": "issue:30064", "accept": false, - "reason": "Early stopping metric resolution and void segmentation map processing are unrelated." + "reason": "Early stopping/eval metric handling has nothing to do with void segmentation map processing." }, { "left": "issue:29159", "right": "issue:36032", "accept": false, - "reason": "Slow/fast tokenizer inconsistency and T5Tokenizer add_special_tokens conflict are different tokenizer failures." + "reason": "Tokenizer inconsistency between slow/fast and T5 add_special_tokens conflict are distinct issues." }, { "left": "issue:43550", "right": "issue:43653", "accept": false, - "reason": "Bamba torch.compile+SDPA failure is unrelated to BigBirdTokenizer mask-token registration." + "reason": "Bamba torch.compile+SDPA failure and BigBird mask-token registration bug are unrelated." }, { "left": "issue:43606", "right": "issue:43720", "accept": false, - "reason": "Both touch loading/offload, but one is a CPU-offload device mismatch and the other is packed-weight unpacking during accelerate loading." + "reason": "Both involve loading, but Bark CPU offload mismatch and BitNet packed-weight unpacking are different fixes." }, { "left": "issue:30064", "right": "issue:34634", "accept": false, - "reason": "Void segmentation maps in image processing and BarkProcessor voice_preset are unrelated." + "reason": "ImageProcessor void segmentation-map handling and BarkProcessor voice_preset are unrelated processor bugs." }, { - "left": "issue:44530", - "right": "issue:46291", + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer on serve is unrelated to Qwen3.5 streaming chat/completions failure." + }, + { + "left": "issue:43421", + "right": "issue:45701", + "accept": false, + "reason": "Runtime special-token/post-processor update behavior is not the same as a generic tokenization-change report." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Qwen3.5 visual-encoder save_pretrained keys and Sam3 text-encoder weights loading are different model-specific serialization bugs." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategies and outdated docs about removed pipelines are both docs-related, but not the same issue." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Different tokenizer/model bugs: Bark CPU offload mismatch vs BigBird special-token registration." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and AutoModelForCausalLM text_config unwrapping are separate Qwen-related issues." + }, + { + "left": "issue:41720", + "right": "issue:43322", "accept": false, - "reason": "PagedAttentionCache group-type crash on Qwen3.5 has no overlap with kernels import metadata requirements." + "reason": "Qwen3 auto device mapping and Llava Next loading segfault are not the same failure." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "FineGrainedFP8Config failure and Llava Next segfault are unrelated model-loading bugs." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Same general area of model serialization/loading, but different models, different attributes, and different root causes." } ] }, @@ -4820,29 +4780,27 @@ "nodes": [ "issue:33666", "issue:34567", + "issue:34634", "issue:35707", "issue:39692", "issue:41720", "issue:42490", "issue:42886", - "issue:42915", - "issue:43322", "issue:43329", "issue:43381", "issue:43388", "issue:43421", + "issue:43525", "issue:43531", - "issue:43606", "issue:43653", "issue:43746", "issue:43844", "issue:43873", "issue:43927", "issue:44062", + "issue:44186", "issue:44368", "issue:44451", - "issue:44492", - "issue:44509", "issue:44514", "issue:44545", "issue:44589", @@ -4856,25 +4814,19 @@ "issue:45042", "issue:45059", "issue:45072", + "issue:45081", + "issue:45356", "issue:45357", - "issue:45406", - "issue:45464", "issue:45588", - "issue:45701", + "issue:45663", "issue:45704", "issue:45759", "issue:45859", "issue:45923", - "issue:46402" + "issue:46032", + "issue:46463" ], "soft_pairs": [ - "issue:45406|issue:45464", - "issue:43421|issue:45701", - "issue:42915|issue:43322", - "issue:45357|issue:46402", - "issue:44492|issue:44509", - "issue:43606|issue:43653", - "issue:43531|issue:45759", "issue:34567|issue:35707", "issue:43388|issue:45059", "issue:43873|issue:45704", @@ -4895,210 +4847,238 @@ "issue:43381|issue:43844", "issue:42886|issue:44843", "issue:45704|issue:45923", - "issue:44918|issue:44936" + "issue:44918|issue:44936", + "issue:46032|issue:46463", + "issue:43421|issue:43653", + "issue:45704|issue:45859", + "issue:45663|issue:45923", + "issue:34634|issue:41720", + "issue:43525|issue:43531", + "issue:43421|issue:44186", + "issue:43381|issue:45704", + "issue:39692|issue:43873", + "issue:45081|issue:45356" ], - "prepared_review_unit_hash": "a59e3c91807e8ac22e845c6ac45f853c612035001729169f145622e0d7fe4483", + "prepared_review_unit_hash": "1ad4db2b9f3557b752ce8c6e628ac0b7df5b743cf7f00b808fd497937fd6196e", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36318, - "estimated_input_tokens": 9080, - "estimated_eval_tokens": 18416 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36897, + "estimated_input_tokens": 9225, + "estimated_eval_tokens": 18706 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly unrelated issues with one clear duplicate pair: the two `Qwen2_5_VLProcessor.apply_chat_template` crash reports. No PRs are present.", - "confidence": 0.92, - "canonical_issue_reason": "issue:44545 is the best canonical target: it describes the same `Qwen2_5_VLProcessor.apply_chat_template` batched-input `padding=False` crash as 44514, with slightly more activity/inbound refs and a cleaner title.", + "summary": "This cluster is mostly heterogeneous: training, tokenizer, multimodal processor, offload, and CI bugs that do not share a single underlying fix. The only clear duplicate pair is the two Qwen2.5-VL `apply_chat_template(..., padding=False)` reports.", + "confidence": 0.93, + "canonical_issue_reason": "Use issue:44514 as the canonical issue for the only true duplicate pair; issue:44545 is the same Qwen2_5_VLProcessor batched `apply_chat_template` crash with `padding=False`, just phrased slightly differently.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44545 is the strongest representative issue in this set because it is a concrete, reproducible bug report and the exact duplicate of 44514.", + "best_issue_reason": "issue:44514 is the best issue representative here because it is a crisp, reproducible bug report and has an exact duplicate (issue:44545).", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:45406", - "right": "issue:45464", - "accept": false, - "reason": "Different failures in different paths: Gemma4Processor `_tokenizer` attribute crash vs Qwen3.5 streaming chat/completions issue." - }, - { - "left": "issue:43421", - "right": "issue:45701", - "accept": false, - "reason": "Tokenizers post-processor update behavior is unrelated to a broad tokenization regression claim." - }, - { - "left": "issue:42915", - "right": "issue:43322", - "accept": false, - "reason": "Different models and symptoms: Qwen3Moe FP8 config failure vs Llava Next segmentation fault." - }, - { - "left": "issue:45357", - "right": "issue:46402", - "accept": false, - "reason": "Both involve Qwen3.5/SAM3-ish model loading, but they are distinct bugs with different failing code paths." - }, - { - "left": "issue:44492", - "right": "issue:44509", - "accept": false, - "reason": "A typo in cache strategy docs is not the same as stale pipeline-task documentation." - }, - { - "left": "issue:43606", - "right": "issue:43653", - "accept": false, - "reason": "CPU offload device mismatch and tokenizer mask-token registration are unrelated." - }, - { - "left": "issue:43531", - "right": "issue:45759", - "accept": false, - "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to `from_config` text_config unwrapping." - }, { "left": "issue:34567", "right": "issue:35707", "accept": false, - "reason": "TrainerState token counting and progressive generation with embeddings/past KV are unrelated bugs." + "reason": "Different failures: TrainerState token counting vs progressive generation with `inputs_embeds`/`past_key_values`; no shared code-path bug." }, { "left": "issue:43388", "right": "issue:45059", "accept": false, - "reason": "Metric gathering last-batch truncation is unrelated to SAM3 text+boxes behavior." + "reason": "Unrelated subsystems: metric gathering/label truncation vs SAM3 text+box prompt behavior." }, { "left": "issue:43873", "right": "issue:45704", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to T5's Apex RMSNorm memory leak." + "reason": "Quantization/offloading behavior vs T5 apex RMSNorm memory leak; different bug classes and fixes." }, { "left": "issue:42490", "right": "issue:43329", "accept": false, - "reason": "A trailing-slash import warning is unrelated to the multimodal token-counting video-branch bug." + "reason": "Local-path slash import warning is unrelated to the multimodal token counting video-branch NameError." }, { "left": "issue:44451", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer loading failures, but they affect different models and likely different loader paths." + "reason": "Both are tokenizer loading regressions, but for different models and different failure modes." }, { "left": "issue:43531", "right": "issue:44589", "accept": false, - "reason": "Qwen3-MoE sliding-window handling is unrelated to Float8 storage lookup failures." + "reason": "Qwen3-MoE sliding_window issue is unrelated to the Float8 storage type error." }, { "left": "issue:33666", "right": "issue:35707", "accept": false, - "reason": "Multi-GPU Qwen2-VL training and progressive generation are different problems." + "reason": "Multi-GPU training request vs generation correctness bug; not the same underlying issue." }, { "left": "issue:44898", "right": "issue:45072", "accept": false, - "reason": "Perceiver image-size interpolation and bfloat16 dtype mismatch are unrelated." + "reason": "Perceiver image-resolution failure and dtype mismatch in SwitchTransformers/TimmWrapperModel are different code paths." }, { "left": "issue:44514", "right": "issue:44545", "accept": true, - "reason": "Same exact `Qwen2_5_VLProcessor.apply_chat_template` batched-input `padding=False` crash; title-only variation." + "reason": "Same bug and same code path: `Qwen2_5_VLProcessor.apply_chat_template` crashes on batched input when `padding=False`." }, { "left": "issue:45759", "right": "issue:45859", "accept": false, - "reason": "Composite multimodal config unwrapping and CPU-offload rotary embedding compatibility are different code paths." + "reason": "Composite multimodal config unwrapping vs CPU offload incompatibility in Qwen3.5 MoE rotary embedding; not mergeable as one fix." }, { "left": "issue:41720", "right": "issue:43746", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and GraniteSpeech PEFT local checkpoint loading are unrelated." + "reason": "Qwen3 auto device mapping CUDA assert and GraniteSpeech local PEFT checkpoint loading are unrelated." }, { "left": "issue:45588", "right": "issue:45759", "accept": false, - "reason": "Flash-attention `s_aux=None` crash is unrelated to multimodal config unwrapping." + "reason": "Flash-attention `s_aux=None` crash is unrelated to Qwen multimodal config handling." }, { "left": "issue:44368", "right": "issue:45357", "accept": false, - "reason": "A config warning about `tie_word_embeddings` is not the same as incorrect visual-encoder keys in `save_pretrained`." + "reason": "Warning about `tie_word_embeddings` during LoRA fine-tuning is not the same as saving incorrect visual encoder keys." }, { "left": "issue:44062", "right": "issue:44451", "accept": false, - "reason": "AddedToken special-argument conflict is unrelated to model loading for ScandiBERT." + "reason": "`AddedToken(..., special=...)` construction error is unrelated to the ScandiBERT loading failure." }, { "left": "issue:43421", - "right": "issue:43927", + "right": "issue:43531", "accept": false, - "reason": "Runtime post-processor updates and save/load custom token ID loss are different tokenizer issues." + "reason": "Runtime special-token post-processor update issue is unrelated to Qwen3-MoE `sliding_window` behavior." }, { "left": "issue:44987", "right": "issue:45042", "accept": false, - "reason": "Loading a specific model and PIL image processors requiring torchvision are unrelated." + "reason": "Loading a specific model in transformers>=5.1.0 and PIL backend processors requiring torchvision are separate regressions." }, { "left": "issue:39692", "right": "issue:44610", "accept": false, - "reason": "SigLIP2 doc example errors and OmDet-Turbo processor/model size mismatch are separate documentation/model issues." + "reason": "SigLIP2 docs example errors and OmDet-Turbo processor/model resolution mismatch are different issues." }, { "left": "issue:43381", "right": "issue:43844", "accept": false, - "reason": "Gradient checkpointing eval-mode restriction and ZeRO-3 gradient growth are different training behaviors." + "reason": "Gradient checkpointing eval-mode restriction is unrelated to DeepSpeed ZeRO-3 gradient growth." }, { "left": "issue:42886", "right": "issue:44843", "accept": false, - "reason": "Offline tokenizer cache loading and unconditional `model_info()` in `_patch_mistral_regex` are closely related to offline mode, but they are different bugs and not clearly the same code-path fix." + "reason": "Tokenizer cache offline failure and unconditional `model_info()` call in `_patch_mistral_regex` are distinct offline-loading bugs." }, { "left": "issue:45704", "right": "issue:45923", "accept": false, - "reason": "Apex RMSNorm memory leak is unrelated to a missing `supports_gradient_checkpointing` flag." + "reason": "T5 apex RMSNorm leak is unrelated to missing `supports_gradient_checkpointing` on Nemotron trust_remote_code." }, { "left": "issue:44918", "right": "issue:44936", "accept": false, - "reason": "Unpacking Qwen3.5 input embeddings in TRL SFT is unrelated to `trainer.evaluate()` failing after `trainer.train()`." + "reason": "SFT embedding unpacking failure is not the same as `trainer.evaluate()` failing after `trainer.train()`." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Mamba2Mixer `use_cache` correctness bug and T5 decoder-input error message bug are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Tokenizer backend post-processor update bug is unrelated to BigBirdTokenizer special-token registration/decoding." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "Memory leak in T5 RMSNorm is unrelated to Qwen3.5 MoE CPU-offload incompatibility." + }, + { + "left": "issue:45663", + "right": "issue:45923", + "accept": false, + "reason": "Gemma-4 FSDP2 shared_kv_states bug is unrelated to a missing class flag on Nemotron." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice preset parsing and Qwen3 auto device mapping CUDA assert are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Llama4Config missing `pad_token_id` and Qwen3-MoE `sliding_window` are different configuration bugs." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "TokenizersBackend special-token update bug is unrelated to LayoutLMv2 NER/padding crash." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode issue and T5 memory leak are separate problems." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "SigLIP2 documentation errors are unrelated to quantization/offloading behavior." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenizer regex/codec handling, but the concrete failures differ: crash in `_patch_mistral_regex` vs Kimi-K2.5 regression and misleading warning." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: the only clearly related pair is 44514/44545, whose titles are effectively identical, and the other soft-edge rejections stay conservative. No PRs are present in the packet." + "feedback": "The summary is grounded in the packet and stays conservative. The duplicate call on issue:44514 / issue:44545 is well supported by the titles, and the other soft-edge rejections do not overclaim sameness based on the provided evidence." }, "evaluator_used": true, "retried": false, @@ -5120,27 +5100,27 @@ "cluster_id": "cluster-10000-459", "nodes": [ "issue:30064", - "issue:34634", "issue:36010", - "issue:39692", - "issue:41720", + "issue:36331", "issue:42915", "issue:43295", - "issue:43381", - "issue:43421", - "issue:43525", + "issue:43335", + "issue:43388", "issue:43531", "issue:43540", "issue:43550", "issue:43596", "issue:43618", + "issue:43643", "issue:43653", "issue:43697", + "issue:43716", "issue:43720", "issue:43746", + "issue:43756", "issue:43782", "issue:43819", - "issue:43873", + "issue:43950", "issue:43994", "issue:44008", "issue:44062", @@ -5151,34 +5131,23 @@ "issue:44625", "issue:44661", "issue:44671", + "issue:44843", + "issue:44877", "issue:44898", "issue:45072", "issue:45081", + "issue:45183", "issue:45356", "issue:45381", "issue:45479", - "issue:45663", - "issue:45704", "issue:45834", - "issue:45859", "issue:45907", - "issue:45923", + "issue:45987", "issue:46032", "issue:46050", - "issue:46294", - "issue:46463" + "issue:46294" ], "soft_pairs": [ - "issue:46032|issue:46463", - "issue:43421|issue:43653", - "issue:45704|issue:45859", - "issue:45663|issue:45923", - "issue:34634|issue:41720", - "issue:43525|issue:43531", - "issue:43421|issue:44186", - "issue:43381|issue:45704", - "issue:39692|issue:43873", - "issue:45081|issue:45356", "issue:43596|issue:44220", "issue:45356|issue:45907", "issue:43540|issue:46294", @@ -5199,222 +5168,225 @@ "issue:43540|issue:44008", "issue:44186|issue:44898", "issue:44062|issue:44488", - "issue:43653|issue:43819" + "issue:43653|issue:43819", + "issue:45183|issue:45479", + "issue:43531|issue:43643", + "issue:43335|issue:43756", + "issue:43388|issue:46032", + "issue:43643|issue:44877", + "issue:43716|issue:44008", + "issue:44843|issue:45081", + "issue:30064|issue:36331", + "issue:43950|issue:45987" ], - "prepared_review_unit_hash": "937266fca631b655b269937aa864ccf1e1ed898df8c3bb110e67e52f134ac32d", + "prepared_review_unit_hash": "52dd0dd85f55d8e52129119895830228d235086b37f631b3e704dca1fb695684", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37243, - "estimated_input_tokens": 9311, - "estimated_eval_tokens": 18878 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36628, + "estimated_input_tokens": 9157, + "estimated_eval_tokens": 18570 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose mix of unrelated Transformers bug reports and feature requests. The soft pairs share broad subsystems at most, but none look like the same underlying defect or change, so all should stay separate.", - "confidence": 0.88, - "canonical_issue_reason": "Issue 46032 is the strongest standalone bug report: it is open, recent, has a precise incorrect-results description, and clearly documents the failing code path across CPU/GPU paths.", + "summary": "This cluster is mostly noisy: the soft pairs are similar only at a broad topic level, not the same underlying bug/change. No PRs are present, and the closest thing to a theme is a small Mistral/tokenizer-regex subgroup.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45081 is the closest anchor in an otherwise heterogeneous set: it is a concrete, well-referenced tokenizer-regex regression, and several nearby items are at least tokenizer/loading related.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 46032 is the highest-signal issue here overall because it is concrete, reproducible, and actionable, with clear user impact.", + "best_issue_reason": "If a representative issue must be chosen, issue:45081 is the strongest fit because it is specific, actionable, and has more discussion/inbound references than the other near-matches.", "best_pr_reason": null, "soft_edge_verdicts": [ { - "left": "issue:46032", - "right": "issue:46463", + "left": "issue:43596", + "right": "issue:44220", "accept": false, - "reason": "Different bug classes: Mamba cache correctness vs T5 decoder-input error messaging." - }, - { - "left": "issue:43421", - "right": "issue:43653", - "accept": false, - "reason": "Both involve tokenizers, but one is runtime post-processor syncing and the other is BigBird special-token registration." - }, - { - "left": "issue:45704", - "right": "issue:45859", - "accept": false, - "reason": "Unrelated issues: T5 Apex RMSNorm memory leak vs Qwen3 MoE CPU offload incompatibility." - }, - { - "left": "issue:45663", - "right": "issue:45923", - "accept": false, - "reason": "Different model/runtime problems: Gemma FSDP2 shared_kv_state bug vs missing gradient-checkpointing flag." - }, - { - "left": "issue:34634", - "right": "issue:41720", - "accept": false, - "reason": "BarkProcessor voice preset failure and Qwen3 auto-device cuda assert are unrelated." - }, - { - "left": "issue:43525", - "right": "issue:43531", - "accept": false, - "reason": "Missing pad_token_id in Llama4Config is not the same as Qwen3-MoE sliding_window behavior." - }, - { - "left": "issue:43421", - "right": "issue:44186", - "accept": false, - "reason": "Both are tokenizer-related, but one is special-token post-processing and the other is LayoutLMv2 NER/padding crashes." - }, - { - "left": "issue:43381", - "right": "issue:45704", - "accept": false, - "reason": "Gradient checkpointing in eval mode is unrelated to the Apex RMSNorm leak." - }, - { - "left": "issue:39692", - "right": "issue:43873", - "accept": false, - "reason": "Doc-example/model mismatch and quantization failure are not the same as offloading-with-quantization behavior." - }, - { - "left": "issue:45081", - "right": "issue:45356", - "accept": false, - "reason": "Both touch fix_mistral_regex, but one is a backend_tokenizer AttributeError and the other is a Kimi tokenizer regression with codec handling." - }, - { - "left": "issue:43596", - "right": "issue:44220", - "accept": false, - "reason": "Deepspeed zero3/BertModel index error is unrelated to fbank feature extraction." + "reason": "Different subsystems and failure modes: DeepSpeed/BERT init index error vs audio feature extraction." }, { "left": "issue:45356", "right": "issue:45907", "accept": false, - "reason": "Tokenizer regression vs list_repo_files revision-forwarding bug are unrelated." + "reason": "Unrelated bugs: tokenizer codec/regex regression vs Hub list_repo_files revision forwarding." }, { "left": "issue:43540", "right": "issue:46294", "accept": false, - "reason": "Both are video-input issues, but they affect different Qwen variants and different failure modes." + "reason": "Both are video-related, but they target different Qwen models and different code paths/errors." }, { "left": "issue:44625", "right": "issue:45479", "accept": false, - "reason": "num_labels propagation in Qwen3.5 is unrelated to the single-label zero-loss classification bug." + "reason": "One is Qwen3.5 config propagation; the other is a generic sequence-classification loss issue." }, { "left": "issue:43994", "right": "issue:44671", "accept": false, - "reason": "Different models and symptoms: SigLIP2 incorrect outputs vs CamemBERT masked-LM prediction regression." + "reason": "Different models and symptoms: SigLIP2 pipeline outputs vs CamemBERT masked-LM predictions." }, { "left": "issue:43782", "right": "issue:45381", "accept": false, - "reason": "Qwen3VL weight_only loading and Qwen2.5-VL video position ids are different problems." + "reason": "Different models and bugs: Qwen3VL weight_only load failure vs Qwen2.5-VL video position-id mismatch." }, { "left": "issue:30064", "right": "issue:36010", "accept": false, - "reason": "ImageProcessor segmentation-map handling is unrelated to the GenerationMixin import error." + "reason": "Completely unrelated: image processor void segmentation maps vs GenerationMixin import error." }, { "left": "issue:43720", "right": "issue:43819", "accept": false, - "reason": "Packed-weight loading bug and DAC from_latents/STE mismatch do not share the same code path." + "reason": "Different code paths: packed-weight loading in BitNet vs DAC latent/STE behavior." }, { "left": "issue:43653", "right": "issue:43720", "accept": false, - "reason": "BigBird special-token registration and BitNet packed-weight unpacking are unrelated." + "reason": "Tokenizer special-token registration bug vs BitNet packed-weight loading bug; not the same issue." }, { "left": "issue:44220", "right": "issue:44661", "accept": false, - "reason": "Audio fbank feature extraction and add-new-model-like tokenizer mapping failures are different subsystems." + "reason": "Audio feature extraction issue vs add-new-model-like/tokenizer mapping failure." }, { "left": "issue:44077", "right": "issue:44625", "accept": false, - "reason": "patchtsmixer post_init policy and Qwen3.5 num_labels propagation are unrelated." + "reason": "PatchTST post_init policy vs Qwen3.5 num_labels propagation; unrelated." }, { "left": "issue:43746", "right": "issue:46050", "accept": false, - "reason": "PEFT local-checkpoint loading and sinq save/load quantization breakage are different issues." + "reason": "PEFT/local checkpoint loading vs quantized save/load corruption; different problems." }, { "left": "issue:43618", "right": "issue:43697", "accept": false, - "reason": "CLIP attentions assignment and RTDetrV2 output changes are not the same bug." + "reason": "CLIP attentions assignment regression vs RTDetrV2 output differences in v5." }, { "left": "issue:42915", "right": "issue:43746", "accept": false, - "reason": "Qwen3Moe FP8 config failure and GraniteSpeech PEFT loading are unrelated." + "reason": "Different models and failure points: Qwen3Moe FP8 config vs GraniteSpeech PEFT loading." }, { "left": "issue:43550", "right": "issue:45072", "accept": false, - "reason": "torch.compile SDPA failure is not the same as bfloat16 dtype mismatch." + "reason": "Both touch performance/dtype areas, but the concrete bugs are different models and different failures." }, { "left": "issue:43295", "right": "issue:45834", "accept": false, - "reason": "Processor/tokenizer regression and Kosmos2.5 OCR index error are unrelated." + "reason": "Processor/tokenizer regression vs Kosmos2.5 long-OCR index error; not the same bug." }, { "left": "issue:43550", "right": "issue:43819", "accept": false, - "reason": "Different failures: compile/SDPA vs DAC from_latents reconstruction mismatch." + "reason": "Unrelated issues: SDPA/torch.compile failure vs DAC latent decoding mismatch." }, { "left": "issue:43540", "right": "issue:44008", "accept": false, - "reason": "Video-input processing error and Gemma3n variable collision are unrelated." + "reason": "Both are multimodal, but one is Qwen3Omni video input handling and the other is a Gemma3n variable-name collision." }, { "left": "issue:44186", "right": "issue:44898", "accept": false, - "reason": "Tokenizer crash on NER/padding and Perceiver image-resolution failure are different bugs." + "reason": "Tokenizer padding/NER crash vs Perceiver image-res interpolation failure; unrelated." }, { "left": "issue:44062", "right": "issue:44488", "accept": false, - "reason": "AddedToken constructor error and sleng-bert loading failure are not the same defect." + "reason": "AddedToken keyword conflict vs loading a specific tokenizer model; both are tokenizer-related but not the same defect." }, { "left": "issue:43653", "right": "issue:43819", "accept": false, - "reason": "BigBird special-token registration and DAC STE mismatch are unrelated." + "reason": "Special-token registration bug vs DAC STE mismatch; no shared code-path." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Validation/error-message bug vs degenerate zero-loss behavior; different bugs and fixes." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding-window bug vs AutoConfig trust_remote_code field loss; unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug vs Smollm3 RoPE-layer dropping; different model/config issues." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Metrics gathering label truncation vs Mamba2 cache correctness; unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both are config-loading issues, but they concern different mechanisms and models." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Different multimodal paths and different error types: image dtype mismatch vs audio_mel_mask collision." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Same helper function, but different concrete failures: offline model_info call vs backend_tokenizer AttributeError." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image segmentation processing vs custom trainer loss signature; unrelated." + }, + { + "left": "issue:43950", + "right": "issue:45987", + "accept": false, + "reason": "Both involve from_pretrained/save behavior, but one is buffer restoration corruption and the other is seq_length return type." } ] }, @@ -5438,64 +5410,56 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:30064", + "issue:28218", + "issue:29159", "issue:34634", "issue:36010", - "issue:36331", "issue:39692", + "issue:42175", "issue:43322", - "issue:43335", - "issue:43388", "issue:43421", - "issue:43531", + "issue:43452", "issue:43540", "issue:43577", + "issue:43582", "issue:43606", - "issue:43643", - "issue:43716", - "issue:43756", + "issue:43720", "issue:43782", "issue:43818", "issue:43825", - "issue:43950", + "issue:43827", "issue:43994", - "issue:44008", "issue:44186", "issue:44442", "issue:44448", "issue:44479", "issue:44485", "issue:44509", + "issue:44561", + "issue:44610", "issue:44671", + "issue:44805", "issue:44811", "issue:44821", - "issue:44843", - "issue:44877", + "issue:44857", + "issue:44936", "issue:44977", + "issue:44991", + "issue:45020", "issue:45072", "issue:45081", "issue:45084", - "issue:45183", "issue:45381", - "issue:45479", "issue:45584", "issue:45685", "issue:45834", "issue:45987", - "issue:46032", + "issue:46192", + "issue:46291", "issue:46294", "issue:46463" ], "soft_pairs": [ - "issue:45183|issue:45479", - "issue:43531|issue:43643", - "issue:43335|issue:43756", - "issue:43388|issue:46032", - "issue:43643|issue:44877", - "issue:43716|issue:44008", - "issue:44843|issue:45081", - "issue:30064|issue:36331", - "issue:43950|issue:45987", "issue:44821|issue:45084", "issue:44448|issue:44671", "issue:39692|issue:45834", @@ -5510,194 +5474,227 @@ "issue:46294|issue:46463", "issue:43322|issue:45685", "issue:43825|issue:43994", + "issue:44485|issue:44509", "issue:34634|issue:36010", - "issue:44485|issue:44509" + "issue:44561|issue:45020", + "issue:44991|issue:46291", + "issue:45987|issue:46192", + "issue:43452|issue:44991", + "issue:44805|issue:44936", + "issue:28218|issue:29159", + "issue:44610|issue:45072", + "issue:43720|issue:45072", + "issue:43577|issue:43582", + "issue:43825|issue:43827", + "issue:42175|issue:45081", + "issue:43582|issue:44857" ], - "prepared_review_unit_hash": "442e473d8494ade2f15614784f242324de9868ae64bb26a98894102f066f8f47", + "prepared_review_unit_hash": "9cf7624fafd615ec615959d2c21fec7c6b9445b14d4ae31dbb8e3ac0b75f3fed", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 25, - "serialized_chars": 35073, - "estimated_input_tokens": 8769, - "estimated_eval_tokens": 17794 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36187, + "estimated_input_tokens": 9047, + "estimated_eval_tokens": 18350 }, "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/config/regression bugs that only share broad Transformers similarity. The only near-adjacent pair is the Mistral tokenizer helper issues, but they still describe different failure modes, so no soft duplicates are accepted.", - "confidence": 0.87, - "canonical_issue_reason": "issue:44843 is the most representative issue in the cluster: it has strong discussion activity, an inbound reference, and sits in the only clearly related sub-area (_patch_mistral_regex / tokenizer loading) among the candidates.", + "summary": "This cluster is mostly unrelated issues. The only clear duplicate-like pair is the tokenizer added-token spacing bug and the broader slow/fast tokenizer inconsistency report.", + "confidence": 0.94, + "canonical_issue_reason": "issue:29159 is the broader tokenizer inconsistency report and is the best umbrella for the only plausible duplicate pair with 28218.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44843 is the best overall representative because it is a concrete, well-documented regression with clear user impact and stronger cross-references than the surrounding items.", + "best_issue_reason": "issue:29159 best represents the duplicate pair because it states the general slow-vs-fast tokenizer inconsistency, while 28218 is the narrower symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:45183", - "right": "issue:45479", - "accept": false, - "reason": "Wrong input validation/XOR error vs a zero-loss classification bug; different code paths and symptoms." - }, - { - "left": "issue:43531", - "right": "issue:43643", - "accept": false, - "reason": "Qwen3-MoE sliding-window behavior vs trust_remote_code config field loss; not the same bug." - }, - { - "left": "issue:43335", - "right": "issue:43756", - "accept": false, - "reason": "Sparse-layer creation bug vs RoPE-layer drop bug; both are model-arch issues but unrelated." - }, - { - "left": "issue:43388", - "right": "issue:46032", - "accept": false, - "reason": "Metric label truncation in gather_for_metrics vs incorrect use_cache results in Mamba2Mixer; distinct problems." - }, - { - "left": "issue:43643", - "right": "issue:44877", - "accept": false, - "reason": "Both involve config loading, but one is missing fields from trust_remote_code and the other is strict config rejection; different underlying failures." - }, - { - "left": "issue:43716", - "right": "issue:44008", - "accept": false, - "reason": "Mistral-3 dtype mismatch vs Gemma3n variable-name collision causing AttributeError; unrelated." - }, - { - "left": "issue:44843", - "right": "issue:45081", - "accept": false, - "reason": "Both touch _patch_mistral_regex, but one fails due to an unconditional model_info() call/offline mode and the other due to missing backend_tokenizer; related area, different concrete bugs." - }, - { - "left": "issue:30064", - "right": "issue:36331", - "accept": false, - "reason": "Image segmentation-map processing bug vs Trainer compute_loss kwarg regression; no shared code path." - }, - { - "left": "issue:43950", - "right": "issue:45987", - "accept": false, - "reason": "from_pretrained corrupting non-persistent buffers vs StaticCache seq-length return-type bug; different subsystems." - }, { "left": "issue:44821", "right": "issue:45084", "accept": false, - "reason": "AutoImageProcessor URL-loading failure vs template compilation error; unrelated." + "reason": "Different bugs: image processor loading from URL vs template compilation error." }, { "left": "issue:44448", "right": "issue:44671", "accept": false, - "reason": "Different model regressions (Pegasus output drift vs CamemBERT masked-LM predictions), not a duplicate bug." + "reason": "Both are v5 regressions, but they affect different models and different failure modes." }, { "left": "issue:39692", "right": "issue:45834", "accept": false, - "reason": "SigLIP2 documentation example errors vs Kosmos2.5 OCR index error; docs/example issue vs runtime model bug." + "reason": "Doc-example mistakes vs a runtime OCR index error; not the same bug." }, { "left": "issue:43421", "right": "issue:44811", "accept": false, - "reason": "Tokenizer post-processor update bug vs Whisper batch_decode skip_special_tokens behavior; different components." + "reason": "Both involve token/special-token handling, but one is post-processor updates and the other is batch_decode skip_special_tokens behavior." }, { "left": "issue:43540", "right": "issue:45381", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError vs qwen2.5-vl vision_position_ids regression; different failure modes." + "reason": "Different Qwen video-input regressions with different symptoms and model paths." }, { "left": "issue:45584", "right": "issue:45834", "accept": false, - "reason": "Whisper empty-transcription generation issue vs Kosmos2.5 long OCR input index error; unrelated." + "reason": "Whisper empty-transcription failure is unrelated to Kosmos2.5 OCR indexing." }, { "left": "issue:43577", "right": "issue:43818", "accept": false, - "reason": "Blip2 dtype handling vs Video-LLaVA temporal-attention/weight-sharing issue; not the same bug." + "reason": "Different models and different loading bugs: dtype propagation vs missing temporal attention/weight sharing." }, { "left": "issue:43782", "right": "issue:44977", "accept": false, - "reason": "Qwen3VL weight_only loading error vs Qwen3.5 flash-attention generation bug; different code paths." + "reason": "Qwen3VL weight_only loading error is not the same as Qwen3.5 flash-attention generation failure." }, { "left": "issue:43606", "right": "issue:45072", "accept": false, - "reason": "Bark CPU offload device mismatch vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel; distinct regressions." + "reason": "Different root causes: CPU offload device mismatch vs bfloat16 dtype mismatch." }, { "left": "issue:44479", "right": "issue:45834", "accept": false, - "reason": "Qwen video-input regression vs Kosmos2.5 OCR indexing; no substantive overlap." + "reason": "Unrelated: Qwen video regression vs Kosmos2.5 long OCR input error." }, { "left": "issue:44186", "right": "issue:44442", "accept": false, - "reason": "LayoutLMv2 tokenizer crash on NER/padding vs AutoTokenizer loading FastSpeech2ConformerTokenizer; unrelated." + "reason": "Tokenizer crash on NER/padding is unrelated to AutoTokenizer loading a FastSpeech2 tokenizer." }, { "left": "issue:46294", "right": "issue:46463", "accept": false, - "reason": "Qwen3.5 video chat-template mismatch vs T5 decoder-input validation messaging; different models and failures." + "reason": "Qwen video feature/token mismatch is unrelated to T5 decoder-input validation." }, { "left": "issue:43322", "right": "issue:45685", "accept": false, - "reason": "Llava Next loading segfault vs MPS histogram-not-implemented error in moe; different runtime issues." + "reason": "Llava Next load segfault and MPS histogram issue are unrelated." }, { "left": "issue:43825", "right": "issue:43994", "accept": false, - "reason": "Pipeline docs mentioning removed tasks vs SigLIP2 AutoModel/pipeline nonsensical results; docs cleanup and model behavior are not the same bug." + "reason": "Pipeline error-message text and SigLIP2 output quality are different issues." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE implementation is unrelated to pipeline-task documentation drift." }, { "left": "issue:34634", "right": "issue:36010", "accept": false, - "reason": "BarkProcessor voice_preset bug vs GenerationMixin import error; unrelated." + "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated." }, { - "left": "issue:44485", - "right": "issue:44509", + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Related theme, but one is a specific helper removal and the other is a broad remote_code breakage report." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading failure and kernels import/version requirement failure are unrelated." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "StaticCache seq-length contract bug and return_tensors kwargs handling are unrelated." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both are loading-related, but gguf_file breakage and a specific tokenizer load failure are different bugs." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask-shape IndexError and trainer.evaluate() after train are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": true, + "reason": "Added-token extra spacing matches the broader slow-vs-fast tokenizer inconsistency report." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor input-size mismatch and dtype mismatch are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "Packed-weight loading failure is not the same as a bfloat16 dtype mismatch." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Blip2 dtype propagation and AppleSilicon allocator warmup TypeError are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Both involve v5 pipeline removal, but one is an error message and the other is docs drift." + }, + { + "left": "issue:42175", + "right": "issue:45081", "accept": false, - "reason": "GLM-5 RoPE implementation discussion vs docs still mentioning removed pipeline tasks; different scope and problem." + "reason": "TF backend packaging and Mistral regex patch crash are unrelated." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "AppleSilicon warmup TypeError and CUDA AMP loss crash 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 plausible duplicate-like pair is issue:28218 with issue:29159, and the rest of the soft-edge rejections are generally supported by the issue titles and descriptions. No PR-pair mergeability concerns apply here." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -5714,67 +5711,56 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:28218", - "issue:29159", + "issue:34634", + "issue:36010", + "issue:36331", + "issue:38175", "issue:42175", "issue:43262", "issue:43329", "issue:43404", - "issue:43452", "issue:43526", "issue:43550", - "issue:43577", - "issue:43582", "issue:43638", "issue:43644", + "issue:43645", + "issue:43650", "issue:43653", "issue:43697", "issue:43720", "issue:43742", + "issue:43746", "issue:43749", "issue:43782", "issue:43819", - "issue:43825", - "issue:43827", + "issue:43854", "issue:43906", + "issue:43927", "issue:44060", "issue:44368", - "issue:44561", "issue:44589", "issue:44610", "issue:44671", - "issue:44805", - "issue:44857", + "issue:44811", "issue:44912", "issue:44936", - "issue:44991", + "issue:44977", "issue:45003", "issue:45020", "issue:45042", - "issue:45072", "issue:45081", "issue:45084", "issue:45406", "issue:45538", "issue:45600", "issue:45701", - "issue:45987", - "issue:46192", "issue:46209", - "issue:46285" + "issue:46285", + "issue:46291", + "issue:46421", + "issue:46424" ], "soft_pairs": [ - "issue:44561|issue:45020", - "issue:45987|issue:46192", - "issue:43452|issue:44991", - "issue:44805|issue:44936", - "issue:28218|issue:29159", - "issue:44610|issue:45072", - "issue:43720|issue:45072", - "issue:43577|issue:43582", - "issue:43825|issue:43827", - "issue:42175|issue:45081", - "issue:43582|issue:44857", "issue:43404|issue:43782", "issue:43262|issue:43819", "issue:44671|issue:45701", @@ -5790,208 +5776,216 @@ "issue:44589|issue:45042", "issue:43644|issue:43749", "issue:43550|issue:44912", - "issue:43638|issue:44936" + "issue:43638|issue:44936", + "issue:43746|issue:44811", + "issue:43819|issue:43927", + "issue:34634|issue:36331", + "issue:43650|issue:43906", + "issue:36010|issue:43645", + "issue:38175|issue:44610", + "issue:42175|issue:46291", + "issue:43720|issue:44610", + "issue:46421|issue:46424", + "issue:43854|issue:44977", + "issue:36010|issue:36331" ], - "prepared_review_unit_hash": "0d04f208633fa4ecd6f82c8d8e843e03c1bf088973773a32e185ee08a6dfdd29", + "prepared_review_unit_hash": "f354d86848190bd3ffc31745044d19453ae7fcca485c6193c00063dcdec2f940", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 36079, - "estimated_input_tokens": 9020, - "estimated_eval_tokens": 18296 + "serialized_chars": 35970, + "estimated_input_tokens": 8993, + "estimated_eval_tokens": 18242 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issues were grouped together; only one pair looks plausibly like the same v5 remote-code regression. The rest are distinct bugs, warnings, or unrelated subsystem failures.", - "confidence": 0.66, - "canonical_issue_reason": "Issue 45020 is the broadest and most representative report of the remote_code breakage in recent transformers versions; 44561 is a narrower specific cause within that same regression family.", + "summary": "This set is mostly a grab bag of unrelated Transformers issues. None of the soft-similarity pairs look like true duplicates or the same underlying fix, so I reject all soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "issue:45020 is the broadest concrete regression report in the set, but there is no strong single canonical because the items are largely unrelated.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45020 is the best cluster representative because it describes the general user-visible failure mode and has the broadest scope among the near-duplicate candidates.", + "best_issue_reason": "issue:45020 is the closest thing to an umbrella anchor among these issues, but only weakly so given the heterogeneity of the cluster.", "best_pr_reason": null, - "soft_edge_verdicts": [ - { - "left": "issue:44561", - "right": "issue:45020", - "accept": true, - "reason": "Both describe recent v5 breakage for remote_code/trust_remote_code models; 44561 is a specific root cause, while 45020 is the broader regression report." - }, - { - "left": "issue:45987", - "right": "issue:46192", - "accept": false, - "reason": "Different code paths: one is a StaticCache return-type contract bug, the other ignores return_tensors when text_kwargs is passed." - }, - { - "left": "issue:43452", - "right": "issue:44991", - "accept": false, - "reason": "Both involve loading/tokenizer behavior, but the concrete failures and likely fixes differ (gguf_file handling vs a specific tokenizer load failure)." - }, - { - "left": "issue:44805", - "right": "issue:44936", - "accept": false, - "reason": "One is a mask/index shape mismatch error, the other is an evaluate-after-train failure; no shared underlying bug is evident." - }, - { - "left": "issue:28218", - "right": "issue:29159", - "accept": false, - "reason": "The second is a broad slow-vs-fast tokenizer inconsistency report; the first is a specific added-token spacing symptom, but they are not clearly the same bug." - }, - { - "left": "issue:44610", - "right": "issue:45072", - "accept": false, - "reason": "Unrelated model issues: input-size mismatch for OmDet-Turbo vs dtype mismatch in SwitchTransformers/TimmWrapperModel." - }, - { - "left": "issue:43720", - "right": "issue:45072", - "accept": false, - "reason": "Different failures: packed-weight unpacking during loading vs bfloat16 dtype mismatch." - }, - { - "left": "issue:43577", - "right": "issue:43582", - "accept": false, - "reason": "BLIP2 dtype loading issue and Apple Silicon allocator TypeError are unrelated." - }, - { - "left": "issue:43825", - "right": "issue:43827", - "accept": false, - "reason": "One is an error-message wording bug, the other is stale docs; related theme, but not the same underlying defect." - }, - { - "left": "issue:42175", - "right": "issue:45081", - "accept": false, - "reason": "TensorFlow packaging/backend inclusion and Mistral tokenizer regex patching are unrelated." - }, - { - "left": "issue:43582", - "right": "issue:44857", - "accept": false, - "reason": "Apple Silicon warmup TypeError is unrelated to CUDA AMP loss crashes." - }, + "soft_edge_verdicts": [ { "left": "issue:43404", "right": "issue:43782", "accept": false, - "reason": "Different model families and failures: tied lm_head weights vs weight_only loading error." + "reason": "Different models and different failures: tied lm_head weights vs a weight_only loading error." }, { "left": "issue:43262", "right": "issue:43819", "accept": false, - "reason": "Audio processor sampling-rate default bug and DAC STE mismatch are separate issues." + "reason": "Audio chat-template sampling-rate bug is unrelated to DAC latent reconstruction mismatch." }, { "left": "issue:44671", "right": "issue:45701", "accept": false, - "reason": "CamemBERT masked-LM regression and general tokenization changes are too broad and not the same bug." + "reason": "CamemBERT masked-LM regression is not the same as a generic tokenization-change report." }, { "left": "issue:45003", "right": "issue:46285", "accept": false, - "reason": "Unsafe sys.modules access and CLIP text-model structure inconsistency are unrelated code-paths." + "reason": "Unsafe sys.modules access is unrelated to CLIP structural inconsistency." }, { "left": "issue:43526", "right": "issue:43697", "accept": false, - "reason": "BeitImageProcessorFast label reduction and RTDetrV2 output differences are different problems." + "reason": "BEiT label reduction bug and RTDetr output drift are different code paths and symptoms." }, { "left": "issue:43653", "right": "issue:44912", "accept": false, - "reason": "Tokenizer special-token registration bug and MXFP4 quantization fallback are unrelated." + "reason": "BigBirdTokenizer decode/special-token issue is unrelated to MXFP4 quantization loading fallback." }, { "left": "issue:45081", "right": "issue:45406", "accept": false, - "reason": "Both are init/load-time attribute errors, but one is Mistral tokenizer regex patching and the other is Gemma4Processor missing _tokenizer; no shared root cause." + "reason": "Both are load-time processor/tokenizer failures, but different models and different missing attributes/code paths." }, { "left": "issue:44060", "right": "issue:44368", "accept": false, - "reason": "Both mention tied-weights warnings in Qwen models, but one reports an incorrect weight-tie bug while the other is about a configuration warning during LoRA fine-tuning." + "reason": "Both mention tied-weight warnings, but they concern different Qwen variants and different warning behavior." }, { "left": "issue:43906", "right": "issue:46209", "accept": false, - "reason": "An isolated reproduction of an older issue and an FP8 inference failure are not the same bug." + "reason": "A reproduction of an older issue is not the same as an FP8 inference failure on B200." }, { "left": "issue:43742", "right": "issue:45020", "accept": false, - "reason": "MobileLLM loading key error is not the same as the broader remote_code regression." + "reason": "MobileLLM key error and remote_code regressions are different loading problems." }, { "left": "issue:45084", "right": "issue:45538", "accept": false, - "reason": "Template compilation TypeError and CLIPTokenizer model_max_length are unrelated." + "reason": "Template-node compilation failure is unrelated to CLIPTokenizer's model_max_length bug." }, { "left": "issue:43329", "right": "issue:45600", "accept": false, - "reason": "Multimodal token counting bug and auto_mappings stale config references are unrelated." + "reason": "Undefined multimodal-token variables and removed auto-mapping configs are separate regressions." }, { "left": "issue:44589", "right": "issue:45042", "accept": false, - "reason": "Float8 storage lookup error and PIL backend torchvision requirement are unrelated." + "reason": "Float8 storage lookup error is not the same as PIL backend requiring torchvision." }, { "left": "issue:43644", "right": "issue:43749", "accept": false, - "reason": "Non-persistent buffer initialization bug and FSDP CPU RAM efficient loading failure are different issues." + "reason": "Junk non-persistent buffers and broken FSDP CPU RAM loading are different initialization/loading defects." }, { "left": "issue:43550", "right": "issue:44912", "accept": false, - "reason": "torch.compile/SDPA failure and MXFP4 quantization fallback are distinct model/runtime problems." + "reason": "Bamba torch.compile+SDPA failure is unrelated to MXFP4 quantization fallback." }, { "left": "issue:43638", "right": "issue:44936", "accept": false, - "reason": "DeepSpeed Zero3 index error and trainer evaluate-after-train failure do not share a concrete root cause." + "reason": "DeepSpeed ZeRO3 index error during training is not the same as evaluate-after-train failure." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT local-checkpoint loading and Whisper skip_special_tokens decoding are unrelated." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC forward/latent mismatch is unrelated to DiaConfig losing custom token IDs on save/load." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature errors are unrelated." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "An empty placeholder issue is not a duplicate of an isolated reproduction report." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import breakage is unrelated to notebook-defined custom model initialization." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Siglip2 zero-probabilities and OmDet processor size mismatch are different model-specific bugs." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Missing tensorflow in an install extra and kernels import requirements are separate dependency issues." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight loading failure for BitNet is unrelated to OmDet-Turbo image-size mismatch." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both are cache-related, but one is lazy-init recompilation and the other is max_cache_length handling." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "GLM-4.7 unit-test loading and Qwen3.5 flash-attention generation are different model-specific regressions." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "ImportError for GenerationMixin and CustomTrainer.compute_loss keyword mismatch are unrelated." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet, and the one accepted soft-edge pair is conservatively justified by the issue titles: 44561 names a specific v5 remote_code breakage cause, while 45020 is the broader user-facing regression report. The remaining pairs are appropriately treated as non-duplicates based on distinct failure modes." - }, - "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 }, @@ -6008,65 +6002,56 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ + "issue:29159", "issue:34567", - "issue:34634", - "issue:36010", - "issue:36331", - "issue:38175", "issue:39692", "issue:40990", "issue:42490", + "issue:42915", "issue:43257", "issue:43262", "issue:43329", + "issue:43404", + "issue:43425", "issue:43450", + "issue:43452", "issue:43454", + "issue:43525", "issue:43526", "issue:43531", "issue:43577", + "issue:43582", "issue:43606", "issue:43643", - "issue:43645", - "issue:43650", - "issue:43720", - "issue:43746", + "issue:43697", + "issue:43742", "issue:43749", - "issue:43819", - "issue:43854", - "issue:43906", "issue:43927", + "issue:43931", "issue:44291", "issue:44387", "issue:44479", "issue:44554", "issue:44568", - "issue:44610", "issue:44617", "issue:44811", + "issue:44821", + "issue:44849", "issue:44912", - "issue:44977", "issue:45042", + "issue:45127", "issue:45356", + "issue:45357", "issue:45372", "issue:45459", + "issue:45701", "issue:45759", "issue:45803", "issue:45859", - "issue:46326", - "issue:46421", - "issue:46424" + "issue:46291", + "issue:46326" ], "soft_pairs": [ - "issue:43746|issue:44811", - "issue:43819|issue:43927", - "issue:34634|issue:36331", - "issue:43650|issue:43906", - "issue:36010|issue:43645", - "issue:38175|issue:44610", - "issue:43720|issue:44610", - "issue:46421|issue:46424", - "issue:43854|issue:44977", - "issue:36010|issue:36331", "issue:39692|issue:44912", "issue:44387|issue:44617", "issue:43531|issue:45042", @@ -6083,208 +6068,229 @@ "issue:43606|issue:45859", "issue:34567|issue:40990", "issue:43329|issue:44479", - "issue:44568|issue:45356" + "issue:44568|issue:45356", + "issue:42915|issue:43742", + "issue:29159|issue:34567", + "issue:43525|issue:46291", + "issue:43404|issue:45127", + "issue:44291|issue:45042", + "issue:43257|issue:43931", + "issue:43927|issue:44912", + "issue:42915|issue:44821", + "issue:43697|issue:45357", + "issue:43452|issue:43582", + "issue:44849|issue:45701", + "issue:43425|issue:46291" ], - "prepared_review_unit_hash": "8902764df33d2df481318c4a9f5d104ab0d4c60fb1f1a70371adbe2524583d31", + "prepared_review_unit_hash": "b27b6bfaca451adaac5e12953306dd54b4c815fb0abfd123b825f47326d67df6", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 27, - "serialized_chars": 35604, - "estimated_input_tokens": 8901, - "estimated_eval_tokens": 18058 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36724, + "estimated_input_tokens": 9181, + "estimated_eval_tokens": 18618 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most suggested pairs are false positives: they share a broad subsystem name (loading, quantization, processors, cache, offload) but describe different concrete failures. The only plausible duplicate is the Qwen video-path regression pair, where the broader v5.3.0 video input regression likely subsumes the specific undefined-variable bug in `_get_num_multimodal_tokens`.", - "confidence": 0.78, - "canonical_issue_reason": "issue:44479 is the broader, more user-facing report and appears to subsume the specific video-token helper failure described in issue:43329.", + "summary": "All candidate links are false positives: they span unrelated tokenizer, model-loading, quantization, processor, and CPU-offload bugs. No pair looks like the same underlying issue/change, so there is no duplicate cluster to collapse.", + "confidence": 0.98, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 is the best anchor for the duplicate cluster because it covers the wider regression surface across multiple Qwen video models while matching the same code path implied by issue:43329.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43746", - "right": "issue:44811", - "accept": false, - "reason": "Different features and failures: GraniteSpeech PEFT checkpoint loading vs Whisper batch decoding skip_special_tokens." - }, - { - "left": "issue:43819", - "right": "issue:43927", - "accept": false, - "reason": "Both are model-loading bugs, but one is DAC latent reconstruction mismatch and the other is DiaConfig save/load token-id loss." - }, - { - "left": "issue:34634", - "right": "issue:36331", - "accept": false, - "reason": "Unrelated areas: BarkProcessor voice_preset vs CustomTrainer compute_loss keyword regression." - }, - { - "left": "issue:43650", - "right": "issue:43906", - "accept": false, - "reason": "Generic placeholder issue vs isolated reproduction of a different bug; no shared underlying code path is apparent." - }, - { - "left": "issue:36010", - "right": "issue:43645", - "accept": false, - "reason": "Both mention Transformers 5.x import/init breakage, but one is GenerationMixin import removal and the other is custom model initialization in notebooks." - }, - { - "left": "issue:38175", - "right": "issue:44610", - "accept": false, - "reason": "Different models and symptoms: zero probabilities in SigLIP2 vs processor input-size mismatch in OmDet-Turbo." - }, - { - "left": "issue:43720", - "right": "issue:44610", - "accept": false, - "reason": "Accelerate packed-weight unpacking bug vs image processor size mismatch; no same bug indication." - }, - { - "left": "issue:46421", - "right": "issue:46424", - "accept": false, - "reason": "Both involve cache behavior, but one is lazy initialization causing recompilation and the other is max_cache_length being honored only for quantized cache." - }, - { - "left": "issue:43854", - "right": "issue:44977", - "accept": false, - "reason": "Different model families and failure modes: GLM-4.7-Flash load issues vs Qwen3.5 flash-attention generation regression." - }, - { - "left": "issue:36010", - "right": "issue:36331", - "accept": false, - "reason": "Import error for GenerationMixin is unrelated to a custom compute_loss signature mismatch." - }, { "left": "issue:39692", "right": "issue:44912", "accept": false, - "reason": "SigLIP2 documentation example errors are unrelated to git-oss-20b MXFP4 loading fallback." + "reason": "Docs/example breakage in SigLIP2 is unrelated to gpt-oss MXFP4 quantization fallback." }, { "left": "issue:44387", "right": "issue:44617", "accept": false, - "reason": "Both end in OOM, but one is int4 quantization reserved-memory growth and the other is a Sam3Video workload memory issue." + "reason": "Both mention OOM, but one is int4 CUDA reserved-memory growth and the other is Sam3Video-specific CPU/GPU usage." }, { "left": "issue:43531", "right": "issue:45042", "accept": false, - "reason": "Qwen3-MoE sliding_window bug and PIL backend torchvision dependency regression are unrelated." + "reason": "Qwen3-MoE sliding_window logic and PIL backend torchvision dependency are different bugs." }, { "left": "issue:45459", "right": "issue:45803", "accept": false, - "reason": "Both concern exception handling, but they affect different tokenizer/conversion paths and different errors." + "reason": "Different exception-handling paths: protobuf decode masking vs bare except in Fuyu tensor conversion." }, { "left": "issue:43526", "right": "issue:44554", "accept": false, - "reason": "BEiT label reduction bug is unrelated to an MPS attention correctness issue." + "reason": "BeitImageProcessorFast label reduction and MPS attention correctness are unrelated code paths." }, { "left": "issue:43749", "right": "issue:44291", "accept": false, - "reason": "Both involve loading, but one is FSDP CPU RAM efficient loading and the other is init_empty_weights argument compatibility." + "reason": "Both concern loading, but FSDP CPU RAM-efficient loading and init_empty_weights argument handling are distinct failures." }, { "left": "issue:43262", "right": "issue:44811", "accept": false, - "reason": "Audio processor chat-template sampling rate defaulting and Whisper batch_decode special-token handling are different bugs." + "reason": "Audio chat-template sampling-rate defaulting is unrelated to Whisper batch_decode skipping special tokens." }, { "left": "issue:43257", "right": "issue:43329", "accept": false, - "reason": "Qwen3 MoE weight-conversion during accelerate+deepspeed loading is unrelated to multimodal video token counting undefined variables." + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is not the same as undefined video token-count helpers." }, { "left": "issue:43643", "right": "issue:45759", "accept": false, - "reason": "AutoConfig missing fields with trust_remote_code and AutoModelForCausalLM text_config unwrapping are distinct configuration bugs." + "reason": "AutoConfig remote-code field loss and composite Qwen text_config unwrapping are different API bugs." }, { "left": "issue:43454", "right": "issue:43577", "accept": false, - "reason": "AyaVision lm_head weight tying and Blip2 dtype propagation are unrelated." + "reason": "Weight tying in AyaVision/Mistral3 is unrelated to Blip2 dtype propagation." }, { "left": "issue:45372", "right": "issue:46326", "accept": false, - "reason": "Both touch mistral_common, but one is a missing ReasoningEffort import and the other is continue_final_message tokenizer behavior." + "reason": "One is an import failure from mistral_common, the other is continue_final_message behavior in a tokenizer." }, { "left": "issue:42490", "right": "issue:43257", "accept": false, - "reason": "Trailing-slash local-path warning is unrelated to Qwen3 MoE weight conversion/loading." + "reason": "Local-path trailing-slash warning is unrelated to Qwen3 MoE loading under accelerate/deepspeed." }, { "left": "issue:43329", "right": "issue:43450", "accept": false, - "reason": "Both mention video, but one is a token-count helper crash and the other is a batched-shape bug in video processors." + "reason": "Undefined video patch helpers and batched video shape errors are separate video-processing bugs." }, { "left": "issue:43606", "right": "issue:45859", "accept": false, - "reason": "Both are CPU-offload issues, but on different models and code paths; not the same bug." + "reason": "Both involve CPU offload, but they affect different models and code paths." }, { "left": "issue:34567", "right": "issue:40990", "accept": false, - "reason": "TrainerState token counter bug is unrelated to GPT-OSS perplexity on WikiText-2." + "reason": "Trainer token-count tracking does not match a perplexity regression on gpt-oss." }, { "left": "issue:43329", "right": "issue:44479", - "accept": true, - "reason": "Both point to the same Qwen video-input regression; the undefined-variable bug in the video branch plausibly explains the broader v5.3.0 failure across affected models." + "accept": false, + "reason": "A missing helper in multimodal token counting is not the same as the v5.3.0 video input regression." }, { "left": "issue:44568", "right": "issue:45356", "accept": false, - "reason": "Tokenizer special-token regression and Kimi-K2.5 codec/warning regression are different tokenizer bugs." + "reason": "Model-specific tokenizer special-token handling and Kimi codec handling are different regressions." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure on Qwen3 MoE is unrelated to MobileLLM key errors." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer parity issues and TrainerState token counters are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is unrelated to kernels package import/version requirements." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Missing lm_head tying and LoRA-extended-vocabulary collapse are different weight-sync problems." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "init_empty_weights argument compatibility has nothing to do with PIL backend torchvision requirements." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3 MoE loading under accelerate/deepspeed is not the same as Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "DiaConfig save/load token IDs loss and git-oss quantization fallback are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3 MoE FP8 loading and AutoImageProcessor URL loading are unrelated issues." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained visual-key corruption are different bugs." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "gguf_file loading failures and Apple Silicon warmup TypeError are unrelated." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-state output bug and version-dependent tokenization changes are different regressions." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch 2.10 compatibility and kernels v0.15 import requirements are separate dependency issues." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and stays appropriately cautious. The single accepted soft-edge pair is framed as a plausible same-code-path Qwen video regression, not as a proven duplicate, and the remaining pairs are conservatively rejected based on distinct titles/symptoms. No obvious overstatement is present." - }, - "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 }, @@ -6302,65 +6308,55 @@ "cluster_id": "cluster-10000-459", "nodes": [ "issue:24643", + "issue:27957", + "issue:28282", "issue:29159", "issue:33357", "issue:34567", + "issue:36010", "issue:36331", "issue:38175", "issue:41720", + "issue:42175", "issue:42915", "issue:43257", "issue:43317", "issue:43404", - "issue:43452", + "issue:43493", "issue:43540", "issue:43577", - "issue:43582", "issue:43646", "issue:43696", - "issue:43697", "issue:43742", "issue:43749", - "issue:43906", - "issue:43927", - "issue:43931", + "issue:43906", + "issue:43994", "issue:44062", "issue:44291", "issue:44479", "issue:44560", "issue:44568", "issue:44737", - "issue:44821", - "issue:44849", + "issue:44792", "issue:44898", - "issue:44912", + "issue:44964", "issue:44987", "issue:44991", - "issue:45042", "issue:45072", - "issue:45127", + "issue:45081", "issue:45137", - "issue:45357", "issue:45381", "issue:45405", "issue:45488", "issue:45538", - "issue:45701", + "issue:45588", "issue:45907", + "issue:45987", "issue:46463", - "issue:46468" + "issue:46468", + "issue:46491" ], "soft_pairs": [ - "issue:42915|issue:43742", - "issue:29159|issue:34567", - "issue:43404|issue:45127", - "issue:44291|issue:45042", - "issue:43257|issue:43931", - "issue:43927|issue:44912", - "issue:42915|issue:44821", - "issue:43697|issue:45357", - "issue:43452|issue:43582", - "issue:44849|issue:45701", "issue:42915|issue:45538", "issue:44987|issue:45538", "issue:41720|issue:43540", @@ -6381,228 +6377,218 @@ "issue:43906|issue:46468", "issue:34567|issue:36331", "issue:43646|issue:43749", - "issue:43577|issue:45072" + "issue:43577|issue:45072", + "issue:44568|issue:46491", + "issue:44737|issue:45588", + "issue:44062|issue:45081", + "issue:43493|issue:43994", + "issue:27957|issue:28282", + "issue:43404|issue:45987", + "issue:44792|issue:44964", + "issue:36010|issue:42175" ], - "prepared_review_unit_hash": "c691f9ffa20359ce1a6e5e14d4f7433d15ca81a2fab297fbb9f49dd614c23d69", + "prepared_review_unit_hash": "e241ea402059b5927ad5fa6f3990a9da798f97f0e0d30cb6fd821c49157511fb", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36976, - "estimated_input_tokens": 9244, - "estimated_eval_tokens": 18744 + "soft_pair_count": 29, + "serialized_chars": 36510, + "estimated_input_tokens": 9128, + "estimated_eval_tokens": 18512 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is mostly unrelated issues across different models/features. The only plausible duplicate pair is the Qwen video-input regression cluster: issue 44479 appears to be the broader report, and 45381 looks like a narrower manifestation of the same underlying vision-position-id/video-input bug.", - "confidence": 0.89, - "canonical_issue_reason": "Choose issue 44479 as canonical: it describes the broader Qwen video-input regression across multiple model variants, while 45381 is a narrower Qwen2.5-VL-specific report that fits the same underlying bug.", + "summary": "Mostly a heterogeneous cluster of unrelated Transformers bugs, with only a few plausible duplicate groups: Qwen 5.3.0 video-input regressions and SigLIP2 output discrepancies. No PRs are present.", + "confidence": 0.81, + "canonical_issue_reason": "issue:44479 is the broadest representative of the strongest overlap in this set: the Qwen 5.3.0 video-input regression family. It covers the shared multimodal/video code path that also appears in the more specific Qwen reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 44479 is the best representative because it captures the widest affected surface area of the one plausible duplicate cluster and is the clearest anchor for triage.", + "best_issue_reason": "issue:44479 is the best single issue anchor because it is the most umbrella-like, recent, and central to the only clear duplicate cluster (Qwen video-input regressions).", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42915", - "right": "issue:43742", - "accept": false, - "reason": "Both are model-loading failures, but they involve different models and different failure modes (FP8 config vs key error)." - }, - { - "left": "issue:29159", - "right": "issue:34567", - "accept": false, - "reason": "Tokenizer slow/fast inconsistency is unrelated to TrainerState token counting; no shared code-path bug." - }, - { - "left": "issue:43404", - "right": "issue:45127", - "accept": false, - "reason": "Both mention tied weights, but one is a missing lm_head tie in a model class and the other is LoRA merge collapse after vocab extension; different concrete defects." - }, - { - "left": "issue:44291", - "right": "issue:45042", - "accept": false, - "reason": "Unexpected init_empty_weights argument error is unrelated to PIL image processor torchvision requirements." - }, - { - "left": "issue:43257", - "right": "issue:43931", - "accept": false, - "reason": "Different Qwen loading problems: accelerate+deepspeed weight conversion vs VL weight-shape mismatch." - }, - { - "left": "issue:43927", - "right": "issue:44912", - "accept": false, - "reason": "DiaConfig save/load token ID loss is unrelated to MXFP4 quantization fallback behavior." - }, - { - "left": "issue:42915", - "right": "issue:44821", - "accept": false, - "reason": "Different subsystems and errors; one is Qwen3Moe FP8 loading, the other is AutoImageProcessor URL loading." - }, - { - "left": "issue:43697", - "right": "issue:45357", - "accept": false, - "reason": "RTDetrV2 output drift in v5 is not the same bug as Qwen3.5 visual encoder key serialization." - }, - { - "left": "issue:43452", - "right": "issue:43582", - "accept": false, - "reason": "GGUF tokenizer/model loading failure is unrelated to an Apple Silicon caching_allocator_warmup TypeError." - }, - { - "left": "issue:44849", - "right": "issue:45701", - "accept": false, - "reason": "Qwen3.5 hidden-state output bug and tokenization-version drift are different issues." - }, { "left": "issue:42915", "right": "issue:45538", "accept": false, - "reason": "FP8 loading failure and CLIPTokenizer max_length default are unrelated." + "reason": "Different subsystems and failure modes: Qwen3MoE FP8 loading vs CLIPTokenizer max length." }, { "left": "issue:44987", "right": "issue:45538", "accept": false, - "reason": "Model loading failure for physical-intelligence/fast is not the same as CLIPTokenizer max_length behavior." + "reason": "Unrelated: a model-loading failure for physical-intelligence/fast versus a CLIP tokenizer constant issue." }, { "left": "issue:41720", "right": "issue:43540", "accept": false, - "reason": "Both are Qwen-family failures, but one is auto device mapping/cuda assert and the other is video-input processing." + "reason": "Both involve Qwen models, but one is auto device mapping/cuda assert and the other is video-input processing; not the same bug." }, { "left": "issue:43317", "right": "issue:45538", "accept": false, - "reason": "Dequantized model offload loading failure is unrelated to tokenizer max_length defaults." + "reason": "Different failures: dequantized model offload loading versus tokenizer max-length behavior." }, { "left": "issue:33357", "right": "issue:34567", "accept": false, - "reason": "MacOS CLIP bus error is unrelated to TrainerState token accounting." + "reason": "MacOS bus error on CLIP versus TrainerState token counter not updating; unrelated." }, { "left": "issue:45137", "right": "issue:45405", "accept": false, - "reason": "DeepSpeed ZeRO3 deque error and PEFT version bump are unrelated." + "reason": "DeepSpeed ZeRO3 deque error and a PEFT version bump are unrelated." }, { "left": "issue:43540", "right": "issue:44479", - "accept": false, - "reason": "Both concern video inputs, but they target different Qwen models and different failure manifestations; not enough evidence of the same concrete bug." + "accept": true, + "reason": "Same Qwen video-input regression family; 44479 is the umbrella report and 43540 is a specific model symptom." }, { "left": "issue:43540", "right": "issue:44560", - "accept": false, - "reason": "Different Qwen video-input errors on different model families; no clear shared underlying defect." + "accept": true, + "reason": "Both are Qwen multimodal video-input failures in the same regression family, likely the same processing bug." }, { "left": "issue:44568", "right": "issue:45488", "accept": false, - "reason": "Both are tokenizer-related, but one is add_special_tokens/BOS-EOS behavior and the other is tokenizer.json pre-tokenizer override; distinct bugs." + "reason": "Different tokenizer regressions affecting different models and code paths." }, { "left": "issue:44479", "right": "issue:45381", "accept": true, - "reason": "These describe the same Qwen video-input regression; 45381 is a specific Qwen2.5-VL manifestation of the broader regression reported in 44479." + "reason": "Same Qwen VL video regression area; 45381 looks like a specific manifestation of the broader 5.3.0 issue." }, { "left": "issue:43742", "right": "issue:44291", "accept": false, - "reason": "Different loading errors with no shared code path or model behavior." + "reason": "Both are model loading errors, but one is a key error for a specific model and the other is an init_empty_weights API break; not the same bug." }, { "left": "issue:44987", "right": "issue:45907", "accept": false, - "reason": "One is a model-loading failure and the other is an API regression in list_repo_files forwarding revision." + "reason": "Loading a model from the Hub and forwarding list_repo_files revision are unrelated." }, { "left": "issue:24643", "right": "issue:29159", "accept": false, - "reason": "DeepSpeed training RuntimeError and tokenizer inconsistency are unrelated." + "reason": "DeepSpeed training weight-shape error and tokenizer inconsistency are unrelated." }, { "left": "issue:38175", "right": "issue:44898", "accept": false, - "reason": "Siglip2 zero probabilities and Perceiver non-default resolution failure are different vision-model issues." + "reason": "Both are vision-model bugs, but they concern different architectures and symptoms." }, { "left": "issue:44737", "right": "issue:46463", "accept": false, - "reason": "XLNet CPU arange regression is unrelated to T5 decoder-input validation messaging." + "reason": "XLNet CPU placement bug and T5 decoder error-message bug are unrelated." }, { "left": "issue:44062", "right": "issue:44991", "accept": false, - "reason": "AddedToken special-argument collision and est-roberta tokenizer loading failure are not the same bug." + "reason": "Both are tokenizer-loading issues, but the titles point to different root causes and model families." }, { "left": "issue:43257", "right": "issue:43696", "accept": false, - "reason": "Qwen3 MoE weight-conversion failure and GPT-oss OOM are different problems." + "reason": "Qwen3 MoE weight-conversion bug versus GPT-oss OOM; unrelated." }, { "left": "issue:43906", "right": "issue:46468", "accept": false, - "reason": "One is a reproduction of an existing issue; the other is a broken official Nemotron example. No evidence they are the same bug." + "reason": "Different models and different failures; no clear shared code-path bug." }, { "left": "issue:34567", "right": "issue:36331", "accept": false, - "reason": "TrainerState token counting and CustomTrainer compute_loss signature breakage are separate trainer issues." + "reason": "TrainerState token accounting and CustomTrainer compute_loss signature mismatch are distinct trainer issues." }, { "left": "issue:43646", "right": "issue:43749", "accept": false, - "reason": "Custom model init breakage in Transformers 5.0 is unrelated to FSDP CPU RAM efficient loading." + "reason": "Custom model initialization break and FSDP CPU RAM efficient loading are separate regressions." }, { "left": "issue:43577", "right": "issue:45072", "accept": false, - "reason": "Blip2 dtype persistence and SwitchTransformers/TimmWrapperModel bf16 mismatch are different dtype-related bugs, not the same concrete failure." + "reason": "Both mention dtype mismatches, but they affect different model families and likely different code paths." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions for different models." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet positional encoding CPU placement and flash-attention crash on sink-less models are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Both are SigLIP2 correctness regressions: one frames it as a HF-vs-JAX discrepancy, the other as nonsensical outputs on the same model family." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention 2 on XLM-Roberta versus missing PyTorch import are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:45987", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 and StaticCache length return type are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal models and different failures; not the same bug." + }, + { + "left": "issue:36010", + "right": "issue:42175", + "accept": false, + "reason": "GenerationMixin import error and TensorFlow backend packaging are unrelated." } ] }, "evaluator_result": { "accept": true, - "feedback": "The conclusions are grounded in the packet and the soft-edge judgments are conservative overall. The 44479/45381 linkage is supported by the shared Qwen video-input / vision_position_ids context, and the rest of the pair rejections stay appropriately cautious given only title-level evidence." + "feedback": "The summary is well grounded in the packet and stays conservative. The Qwen video-input issues are the clearest overlap, and the SigLIP2 pair is a plausible second cluster. No PRs are present, and the soft-edge verdicts do not overclaim beyond the metadata." }, "evaluator_used": true, "retried": false, @@ -6623,64 +6609,55 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:27957", "issue:28282", "issue:29127", "issue:34689", - "issue:36010", + "issue:41762", "issue:42175", + "issue:42898", "issue:43257", - "issue:43404", "issue:43452", - "issue:43493", "issue:43526", "issue:43531", + "issue:43582", "issue:43643", "issue:43697", + "issue:43720", "issue:43742", "issue:43749", "issue:43756", "issue:43824", "issue:43906", - "issue:43994", - "issue:44062", "issue:44291", "issue:44488", - "issue:44568", "issue:44589", - "issue:44737", + "issue:44625", "issue:44756", - "issue:44792", "issue:44821", + "issue:44877", + "issue:44898", "issue:44913", - "issue:44964", + "issue:44936", + "issue:44977", "issue:45042", "issue:45081", + "issue:45200", "issue:45216", + "issue:45325", "issue:45357", "issue:45375", "issue:45381", "issue:45405", "issue:45479", - "issue:45588", "issue:45600", "issue:45823", "issue:45920", - "issue:45987", "issue:46121", "issue:46399", "issue:46463", "issue:46491" ], "soft_pairs": [ - "issue:44568|issue:46491", - "issue:44737|issue:45588", - "issue:44062|issue:45081", - "issue:43493|issue:43994", - "issue:27957|issue:28282", - "issue:43404|issue:45987", - "issue:44792|issue:44964", - "issue:36010|issue:42175", "issue:44291|issue:44821", "issue:28282|issue:34689", "issue:42175|issue:43824", @@ -6699,198 +6676,199 @@ "issue:44913|issue:46121", "issue:43526|issue:45479", "issue:43643|issue:45375", - "issue:43697|issue:45216" + "issue:43697|issue:45216", + "issue:43720|issue:44898", + "issue:45200|issue:46463", + "issue:44625|issue:44877", + "issue:43756|issue:45325", + "issue:42898|issue:46491", + "issue:43582|issue:44977", + "issue:41762|issue:44936" ], - "prepared_review_unit_hash": "bbb37cff7391431d48e7b9bf95046281e26274fe95cee16ecd17631b49e33de8", + "prepared_review_unit_hash": "0b09a0db05cc6b9f483a7c42be4182a76e87abab3620bf948276207888f69ab3", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 27, - "serialized_chars": 36194, - "estimated_input_tokens": 9049, - "estimated_eval_tokens": 18354 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35265, + "estimated_input_tokens": 8817, + "estimated_eval_tokens": 17890 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a heterogeneous collection of Transformer bugs and regressions across tokenizers, model loading, config handling, attention, and multimodal paths. The soft pairs mostly share surface-level subsystem similarity, but they do not appear to describe the same underlying bug/change, so none should be merged as duplicates.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "This cluster is mostly a set of unrelated model-loading, config, tokenizer, and regression bugs; I don\u2019t see any soft pair that looks like the same underlying defect, so all soft edges should be rejected. If a single anchor is needed, issue 45920 is the broadest, most general regression report in the set, but the cluster is not a true duplicate group.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 45920 is the broadest and most central report here: it describes a multi-model tokenizer regression with wide blast radius, whereas the others are narrower, model-specific bugs or unrelated failures.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45920 is the broadest and most central regression report in the cluster, affecting multiple tokenizers/models rather than a single narrow edge case, so it is the closest thing to a representative issue.", + "best_issue_reason": "Issue 45920 is the best representative issue if one must be chosen, because its scope is broad and it reads like a family-level regression rather than a single isolated model bug.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44568", - "right": "issue:46491", - "accept": false, - "reason": "Both are tokenizer round-trip/token handling problems, but one is about add_special_tokens failing to add BOS/EOS for mDeBERTa, while the other is a CodeLlama leading-space encode\u2192decode regression. Different models and different failure modes." - }, - { - "left": "issue:44737", - "right": "issue:45588", - "accept": false, - "reason": "XLNet CPU device placement in relative positional encoding is unrelated to a flash-attention crash on sink-less models; same general runtime area, but not the same bug." - }, - { - "left": "issue:44062", - "right": "issue:45081", - "accept": false, - "reason": "Both involve tokenizer internals, but one is an AddedToken special-argument conflict and the other is a Mistral regex patch crashing on a missing backend_tokenizer attribute. Different code paths and symptoms." - }, - { - "left": "issue:43493", - "right": "issue:43994", - "accept": false, - "reason": "Both concern SigLIP2 behavior, but one is a discrepancy vs the original JAX implementation and the other is nonsensical outputs when used via AutoModel/pipeline. Too broad to be the same concrete bug." - }, - { - "left": "issue:27957", - "right": "issue:28282", - "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and an AutoModel ImportError for missing PyTorch are unrelated issues." - }, - { - "left": "issue:43404", - "right": "issue:45987", - "accept": false, - "reason": "One is a weight-tying bug in Mistral3ForConditionalGeneration; the other is StaticCache.get_seq_length() returning the wrong type. Different components and failure modes." - }, - { - "left": "issue:44792", - "right": "issue:44964", - "accept": false, - "reason": "Both affect model loading for multimodal models, but one is a janus image-generation test failure and the other is a Phi-4-multimodal load regression. No evidence they share the same code defect." - }, - { - "left": "issue:36010", - "right": "issue:42175", - "accept": false, - "reason": "Both are import/backend related, but one is a missing GenerationMixin import and the other is an installation/dependency packaging issue for TensorFlow in a torch-only install. Not the same bug." - }, { "left": "issue:44291", "right": "issue:44821", "accept": false, - "reason": "Unexpected _is_hf_initialized argument during init_empty_weights is a model-construction bug; AutoImageProcessor from URL is an image-processor loading issue. Different paths." + "reason": "Different code paths: init_empty_weights/_is_hf_initialized argument handling vs loading AutoImageProcessor from a URL." }, { "left": "issue:28282", "right": "issue:34689", "accept": false, - "reason": "Missing PyTorch for AutoModel and Llama 3.2 Vision load breakage are distinct; one is environment dependency detection, the other is a model-loading regression." + "reason": "A missing PyTorch import error is not the same as a Llama 3.2 Vision model-loading regression." }, { "left": "issue:42175", "right": "issue:43824", "accept": false, - "reason": "TensorFlow missing from a torch install is packaging-related, whereas Qwen2.5-VL import failure is a model export/import issue. No concrete shared bug." + "reason": "Package/backend dependency mismatch vs missing Qwen2_5_VL import; unrelated failures." }, { "left": "issue:43749", "right": "issue:44756", "accept": false, - "reason": "FSDP CPU RAM efficient loading and disabling mmap on Strix Halo to avoid OOM are both memory-related, but they target different mechanisms and platforms." + "reason": "Both are loading/memory-related, but one is FSDP CPU-efficient loading and the other is mmap/OOM on Strix Halo; not the same bug." }, { "left": "issue:45823", "right": "issue:46463", "accept": false, - "reason": "Gemma4 PLE device mismatch during forward and T5 decoder-input validation error messaging are unrelated model/runtime issues." + "reason": "Gemma4 device_map forward mismatch and T5 decoder input validation are different subsystems and symptoms." }, { "left": "issue:45042", "right": "issue:45081", "accept": false, - "reason": "PIL backend requiring torchvision and Mistral regex patch crashes are different subsystems and failure points." + "reason": "PIL backend dependency regression vs Mistral tokenizer regex crash; no shared underlying defect." }, { "left": "issue:43742", "right": "issue:44589", "accept": false, - "reason": "Loading facebook/MobileLLM-125M with a key error and missing Float8 storage object are separate loading/runtime failures." + "reason": "MobileLLM key-error loading failure vs missing Float8 storage type during deserialization; different causes." }, { "left": "issue:43257", "right": "issue:43756", "accept": false, - "reason": "Qwen3 MoE weights not converted under accelerate+deepspeed and Smollm3 RoPE layer mismatch are unrelated model-specific implementation bugs." + "reason": "Qwen3 MoE weight conversion/loading bug vs Smollm3 RoPE-layer selection bug; not the same issue." }, { "left": "issue:43452", "right": "issue:44488", "accept": false, - "reason": "gguf_file breaking AutoTokenizer/AutoModel loading and failure to load cjvt/sleng-bert are both loading issues, but they affect different APIs and likely different root causes." + "reason": "gguf_file breaks tokenizer/model loading, but cjvt/sleng-bert is a separate model-loading failure." }, { "left": "issue:43906", "right": "issue:45381", "accept": false, - "reason": "An isolated reproduction of an earlier issue and a qwen2.5-vl video position-id regression are not the same concrete bug." + "reason": "One is a reproduction of an older issue; the other is a Qwen2.5-VL video position_ids bug." }, { "left": "issue:45405", "right": "issue:45600", "accept": false, - "reason": "A bumped MIN_PEFT_VERSION that is unreleased and stale auto_mappings references to removed configs are both maintenance regressions, but they are different fixes and different code paths." + "reason": "A version-bump/release-management issue vs stale auto_mappings references; unrelated." }, { "left": "issue:45357", "right": "issue:45920", "accept": false, - "reason": "Qwen3.5 save_pretrained visual-encoder key regression is a checkpoint serialization bug, while issue:45920 is a tokenizer ID regression affecting multiple models. Not the same defect." + "reason": "Qwen3.5 save_pretrained visual-key regression is different from a broad AutoTokenizer token-ID regression." }, { "left": "issue:28282", "right": "issue:29127", "accept": false, - "reason": "AutoModel missing PyTorch and a clearer error for insufficient layoutlmv3 boxes are unrelated." + "reason": "Missing PyTorch import handling is unrelated to LayoutLMv3 box-information error messaging." }, { "left": "issue:43531", "right": "issue:44821", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor loading from URL are different features and failure modes." + "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor URL loading are unrelated." }, { "left": "issue:43756", "right": "issue:46399", "accept": false, - "reason": "Smollm3 RoPE-layer dropping and a missing decorator causing use_kernels=True crashes in Qwen3.5 are separate model/configuration bugs." + "reason": "RoPE-layer dropping in Smollm3 is not the same as a missing decorator causing kernel registration failure." }, { "left": "issue:44913", "right": "issue:46121", "accept": false, - "reason": "GPTNeoX rotary_pct reload regression and convert_rope_params_to_dict rejecting JSON-loaded lists are both RoPE-related, but one is config persistence and the other is validation/type handling." + "reason": "Config reload losing rotary_pct is separate from JSON-loaded list handling in convert_rope_params_to_dict." }, { "left": "issue:43526", "right": "issue:45479", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels truncation and degenerate zero loss in sequence classification are unrelated preprocessing vs training-loss bugs." + "reason": "BeitImageProcessorFast label reduction bug and zero-loss single-label classification bug are different behaviors." }, { "left": "issue:43643", "right": "issue:45375", "accept": false, - "reason": "Missing fields from trust_remote_code AutoConfig returns and Qwen3_5MoeVisionConfig dropping deepstack_visual_indexes are both config-field preservation issues, but for different config classes and mechanisms." + "reason": "trust_remote_code missing fields and strict-dropping deepstack_visual_indexes are different config-serialization issues." }, { "left": "issue:43697", "right": "issue:45216", "accept": false, - "reason": "RTDetrV2 output drift in v5 and Qwen3.5 save_pretrained regression are distinct problems affecting different model families and code paths." + "reason": "RTDetrV2 output drift in v5 is not the same as Qwen3.5 save_pretrained saving incorrect checkpoint keys." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "Packed-weight unpacking during accelerate loading is unrelated to Perceiver non-default resolution failure." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaulting and T5 decoder validation errors are different bugs." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is separate from granite_speech being blocked by strict config loading." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen/Smollm visual-position logic; one is RoPE-layer omission, the other is still-image temporal position_ids scaling." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "Tokenization-space cleanup behavior in v5 and CodeLlama round-trip spacing regression are not the same defect." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "AppleSilicon warmup TypeError vs Qwen3.5 flash-attention generation bug; different execution paths." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO-3 loading IndexError is unrelated to trainer.evaluate() failing after train() due to trainer state." } ] }, @@ -6915,62 +6893,55 @@ "cluster_id": "cluster-10000-459", "nodes": [ "issue:24643", + "issue:28282", "issue:29127", "issue:29159", "issue:33357", + "issue:35141", "issue:38175", - "issue:41762", - "issue:42898", "issue:43295", "issue:43452", "issue:43531", + "issue:43550", "issue:43577", - "issue:43582", + "issue:43643", "issue:43653", "issue:43697", - "issue:43720", "issue:43742", "issue:43756", "issue:43819", + "issue:43927", "issue:44361", "issue:44442", + "issue:44464", "issue:44466", "issue:44484", "issue:44493", "issue:44514", + "issue:44561", "issue:44568", - "issue:44625", "issue:44704", + "issue:44743", "issue:44811", "issue:44843", - "issue:44877", - "issue:44898", "issue:44936", "issue:44977", "issue:44991", "issue:45005", - "issue:45200", - "issue:45325", + "issue:45290", "issue:45356", "issue:45372", "issue:45381", + "issue:45491", "issue:45542", "issue:45823", + "issue:46002", "issue:46050", "issue:46170", "issue:46178", - "issue:46402", - "issue:46463", - "issue:46491" + "issue:46402" ], "soft_pairs": [ - "issue:43720|issue:44898", - "issue:45200|issue:46463", - "issue:44625|issue:44877", - "issue:43756|issue:45325", - "issue:42898|issue:46491", - "issue:43582|issue:44977", - "issue:41762|issue:44936", "issue:44466|issue:45005", "issue:24643|issue:29127", "issue:43452|issue:44843", @@ -6991,204 +6962,225 @@ "issue:44843|issue:44991", "issue:43756|issue:46402", "issue:29159|issue:33357", - "issue:43531|issue:43742" + "issue:43531|issue:43742", + "issue:43653|issue:43927", + "issue:44464|issue:44743", + "issue:44514|issue:45290", + "issue:43550|issue:43927", + "issue:43550|issue:44361", + "issue:44464|issue:44977", + "issue:28282|issue:35141", + "issue:43643|issue:44561", + "issue:45491|issue:46002" ], - "prepared_review_unit_hash": "c7e5425f7b41036b03312bb955147e4b72f36b2bbc1ab69a46fae1a10db73d7b", + "prepared_review_unit_hash": "9421f9274c5c9dbdacfd10bd437ecb8a5a86c9783acfb6b6382608170915cf82", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36402, - "estimated_input_tokens": 9101, - "estimated_eval_tokens": 18458 + "soft_pair_count": 30, + "serialized_chars": 37017, + "estimated_input_tokens": 9255, + "estimated_eval_tokens": 18766 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a heterogeneous set of unrelated bug reports, mostly around tokenizers, processors, model loading, and generation regressions. The soft links are based on broad lexical similarity, not the same underlying defect, so none of the candidate pairs should be merged as duplicates.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "A heterogeneous cluster of mostly unrelated closed issues around tokenizers, model loading, generation, backend imports, and regression edge cases. No soft pair looks like the same underlying bug/change, so all candidate duplicates are rejected. If forced to pick a representative issue, issue:43452 is the broadest loading-path regression in the set.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43452 is the closest cluster anchor because it covers a broad from_pretrained loading-path regression affecting both tokenizer and causal LM loading.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:43452 is the most representative of the cluster\u2019s shared theme of model-loading/tokenizer regressions while remaining a concrete bug report.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43720", - "right": "issue:44898", - "accept": false, - "reason": "Different subsystems and failure modes: BitNet loading/unpacking vs Perceiver positional interpolation." - }, - { - "left": "issue:45200", - "right": "issue:46463", - "accept": false, - "reason": "Gemma 4 multimodal token defaults and T5 decoder-input validation are unrelated bugs." - }, - { - "left": "issue:44625", - "right": "issue:44877", - "accept": false, - "reason": "One is config propagation for Qwen3.5; the other is strict config rejection for granite_speech." - }, - { - "left": "issue:43756", - "right": "issue:45325", - "accept": false, - "reason": "Both mention RoPE, but they affect different models and different logic paths." - }, - { - "left": "issue:42898", - "right": "issue:46491", - "accept": false, - "reason": "Both are tokenizer regressions, but one is a v5 cleanup-setting change and the other is a CodeLlama round-trip spacing bug." - }, - { - "left": "issue:43582", - "right": "issue:44977", - "accept": false, - "reason": "Apple Silicon allocator warmup TypeError is unrelated to Qwen3.5 flash-attention generation." - }, - { - "left": "issue:41762", - "right": "issue:44936", - "accept": false, - "reason": "Gemma3 ZeRO-3 load failure and Trainer evaluate-after-train failure are distinct code paths." - }, { "left": "issue:44466", "right": "issue:45005", "accept": false, - "reason": "Both involve tied weights, but they describe different serialization/loading behaviors and are not the same bug." + "reason": "Both mention tied weights, but one is a device-dependent lm_head serialization inconsistency and the other is a translation-model tied-weights problem; different failures and likely different fixes." }, { "left": "issue:24643", "right": "issue:29127", "accept": false, - "reason": "DeepSpeed training weight-shape error is unrelated to LayoutLMv3 box-info error messaging." + "reason": "DeepSpeed training weight-shape runtime error vs LayoutLMv3 box validation/error-message clarity; unrelated code paths." }, { "left": "issue:43452", "right": "issue:44843", "accept": false, - "reason": "Both affect from_pretrained/tokenizer loading, but one is gguf handling and the other is an offline model_info call in regex patching." + "reason": "Both hit AutoTokenizer/from_pretrained, but one is broken by gguf_file handling and the other by an unconditional model_info() call in a regex patch; different bugs." }, { "left": "issue:44514", "right": "issue:45381", "accept": false, - "reason": "Both are Qwen2.5-VL issues, but one is batched chat template padding and the other is video vision_position_ids." + "reason": "Both involve Qwen2.5-VL, but one is batched apply_chat_template padding behavior and the other is incorrect video vision_position_ids; not the same bug." }, { "left": "issue:43819", "right": "issue:44811", "accept": false, - "reason": "DAC latent reconstruction and Whisper batch_decode token skipping are unrelated." + "reason": "DAC.from_latents vs Whisper batch_decode(skip_special_tokens); no shared underlying defect." }, { "left": "issue:45372", "right": "issue:46170", "accept": false, - "reason": "Gemma 4 processor import regression is unrelated to Windows ROCm torch.distributed import/type errors." + "reason": "Both are import/dependency failures, but they concern different packages, platforms, and missing symbols." }, { "left": "issue:44361", "right": "issue:44442", "accept": false, - "reason": "Different tokenizer classes and different failures: MLuke task AttributeError vs FastSpeech2Conformer loading." + "reason": "Different tokenizers and different failures: MLukeTokenizer AttributeError vs AutoTokenizer failing to load FastSpeech2ConformerTokenizer." }, { "left": "issue:38175", "right": "issue:43295", "accept": false, - "reason": "Siglip2 zero probabilities and processor.tokenizer regression are separate model/processor issues." + "reason": "Siglip2 zero-probability output is unrelated to the processor.tokenizer regression in v4.57.5." }, { "left": "issue:44936", "right": "issue:45823", "accept": false, - "reason": "Trainer lifecycle bug and Gemma4 device-map forward mismatch are not the same defect." + "reason": "trainer.evaluate() after train is a training-loop issue, while Gemma4 PLE device mismatch is a model/device placement bug." }, { "left": "issue:43653", "right": "issue:44361", "accept": false, - "reason": "BigBird special-token registration and MLuke task AttributeError are unrelated tokenizer bugs." + "reason": "Both concern tokenizer behavior, but one is BigBird special-token decode registration and the other is an MLukeTokenizer AttributeError." }, { "left": "issue:44484", "right": "issue:44704", "accept": false, - "reason": "A save_pretrained shard-size question is unrelated to AutoProcessor kwarg forwarding." + "reason": "save_pretrained shard-size question is unrelated to AutoProcessor not forwarding kwargs to cached_file." }, { "left": "issue:43452", "right": "issue:43577", "accept": false, - "reason": "gguf loading breakage and Blip2 dtype not propagating are different load-time bugs." + "reason": "gguf_file loading breakage is unrelated to Blip2 dtype propagation during loading." }, { "left": "issue:45542", "right": "issue:46178", "accept": false, - "reason": "Both concern TensorFlow backends, but one is missing TF installation handling and the other is an unsupported backend mapping entry." + "reason": "Both touch backend selection, but one is undefined tf backend behavior and the other is an invalid tensorflow_text backend mapping." }, { "left": "issue:43295", "right": "issue:45356", "accept": false, - "reason": "Different regressions in different tokenizer/processor code paths and different models." + "reason": "Processor/tokenizer regression and Kimi tokenizer codec/warning regression are different code paths and symptoms." }, { "left": "issue:43577", "right": "issue:43697", "accept": false, - "reason": "Blip2 dtype propagation and RTDetrV2 output changes are unrelated." + "reason": "Blip2 dtype loading and RTDetrV2 output differences are unrelated model bugs." }, { "left": "issue:44493", "right": "issue:46050", "accept": false, - "reason": "Unexpected position-id keys and sinq quantized save/load breakage are distinct issues." + "reason": "Unexpected position-id key warnings are unrelated to NaNs in classification heads." }, { "left": "issue:43653", "right": "issue:44568", "accept": false, - "reason": "BigBird mask-token registration and mDeBERTa special-token addition failure are different tokenizer bugs." + "reason": "Both involve special-token behavior, but one is a BigBird CI/decode issue and the other is mdeberta-v3 not adding BOS/EOS; distinct tokenizer bugs." }, { "left": "issue:44843", "right": "issue:44991", "accept": false, - "reason": "Both are tokenizer-loading regressions, but the offline model_info bug is separate from the est-roberta loading failure." + "reason": "Both are tokenizer-loading regressions, but the offline model_info bug and the est-roberta loading failure are different causes." }, { "left": "issue:43756", "right": "issue:46402", "accept": false, - "reason": "Smollm3 RoPE-layer dropping and Sam3Model text-encoder weights not loading are unrelated." + "reason": "Smollm3 RoPE-layer mismatch and Sam3 text-encoder weights not loaded are unrelated model-specific loading bugs." }, { "left": "issue:29159", "right": "issue:33357", "accept": false, - "reason": "Tokenizer slow/fast inconsistency and CLIP MacOS bus error are unrelated." + "reason": "Slow/fast tokenizer inconsistency is unrelated to a MacOS bus error on a CLIP model." }, { "left": "issue:43531", "right": "issue:43742", "accept": false, - "reason": "Qwen3-MoE sliding_window behavior and MobileLLM key error are different model-loading/runtime problems." + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM key error are different model/configuration bugs." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token registration and DiaConfig custom token-ID persistence are both token-related but not the same failure." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and recurrent-state reset in modular_qwen3_5 are separate generation/cache issues." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both are apply_chat_template crashes, but one is batched padding=False behavior and the other is assistant tool-call messages with no content." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile with SDPA in Bamba is unrelated to DiaConfig token ID loss after save/load." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure is unrelated to MLukeTokenizer AttributeError." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Both mention generation, but chunked compiled forward inconsistency and Qwen3.5 flash-attention generation are different concrete paths." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "Missing PyTorch import for AutoModel is unrelated to post_init reinitializing output embeddings after resize_token_embeddings()." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both involve trust_remote_code regressions, but one loses fields in AutoConfig and the other breaks due to removal of is_torch_fx_available." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both report NaNs, but one is Gemma3 sliding-window attention on GPU and the other is classification-head NaNs on checkout/init; not the same bug." } ] }, @@ -7215,7 +7207,6 @@ "issue:28282", "issue:32090", "issue:33666", - "issue:35141", "issue:36010", "issue:36331", "issue:39692", @@ -7230,11 +7221,10 @@ "issue:43572", "issue:43577", "issue:43606", - "issue:43643", - "issue:43653", "issue:43688", "issue:43720", "issue:43742", + "issue:43749", "issue:43825", "issue:43927", "issue:43994", @@ -7242,35 +7232,28 @@ "issue:44451", "issue:44464", "issue:44466", - "issue:44514", - "issue:44561", + "issue:44479", + "issue:44554", + "issue:44568", "issue:44610", - "issue:44743", + "issue:44625", "issue:44821", "issue:44849", "issue:44964", "issue:44977", - "issue:45290", + "issue:45200", + "issue:45229", + "issue:45357", "issue:45405", "issue:45440", - "issue:45491", "issue:45588", "issue:45823", - "issue:46002", "issue:46050", "issue:46192", - "issue:46209" + "issue:46209", + "issue:46489" ], "soft_pairs": [ - "issue:43653|issue:43927", - "issue:44464|issue:44743", - "issue:44514|issue:45290", - "issue:43550|issue:43927", - "issue:43550|issue:44361", - "issue:44464|issue:44977", - "issue:28282|issue:35141", - "issue:43643|issue:44561", - "issue:45491|issue:46002", "issue:43504|issue:43720", "issue:43504|issue:43550", "issue:43504|issue:44610", @@ -7295,240 +7278,235 @@ "issue:44849|issue:46209", "issue:43493|issue:43825", "issue:44964|issue:45405", - "issue:39692|issue:43493" + "issue:39692|issue:43493", + "issue:44479|issue:44625", + "issue:43577|issue:44554", + "issue:44849|issue:45357", + "issue:44568|issue:46489", + "issue:45200|issue:45229", + "issue:39692|issue:43295", + "issue:43688|issue:43749" ], - "prepared_review_unit_hash": "8dafded6ae124b7b425e406afab41167319b0abef51094c35183e11a96641bda", + "prepared_review_unit_hash": "bfac684d87a90facfdf5dc30e82e03e6c327229fbc0ad2804ea41b789a64dc2a", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 38096, - "estimated_input_tokens": 9524, - "estimated_eval_tokens": 19304 + "soft_pair_count": 32, + "serialized_chars": 37390, + "estimated_input_tokens": 9348, + "estimated_eval_tokens": 18952 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These items are mostly unrelated regressions across loading, generation, tokenizer, and model-architecture paths. None of the soft pairs looks like a true duplicate, so I would not merge any of them; issue 43927 is the cleanest single representative of a concrete, reproducible config/load bug.", - "confidence": 0.32, - "canonical_issue_reason": "Issue 43927 is the most self-contained, reproducible bug in the set: custom token IDs are lost on save/load and then trigger an IndexError during generation.", + "summary": "No soft-edge duplicate pairs look mergeable here; the issues mostly span different models, configs, or code paths, even when the symptom class is similar (loading, generation, device mapping, etc.).", + "confidence": 0.94, + "canonical_issue_reason": "If forced to anchor this heterogeneous set, issue 43452 is the most generic user-facing loading regression, but there is no truly canonical duplicate cluster to collapse around.", "canonical_pr_reason": null, - "best_issue_reason": "Same reason: it has a clear failure mode and a likely actionable code path, making it the best representative issue here.", + "best_issue_reason": "Issue 43452 is the best single representative only in the loose sense that it covers a broad from_pretrained loading failure; otherwise the set is too diverse for a strong global anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43653", - "right": "issue:43927", - "accept": false, - "reason": "Different bugs: tokenizer special-token registration vs config fields being dropped on save/load." - }, - { - "left": "issue:44464", - "right": "issue:44743", - "accept": false, - "reason": "Both affect generation correctness, but they describe different mechanisms and code paths." - }, - { - "left": "issue:44514", - "right": "issue:45290", - "accept": false, - "reason": "Both are chat-template crashes, but the input conditions and failure causes differ." - }, - { - "left": "issue:43550", - "right": "issue:43927", - "accept": false, - "reason": "Unrelated: torch.compile/SDPA failure vs config serialization losing custom token IDs." - }, - { - "left": "issue:43550", - "right": "issue:44361", - "accept": false, - "reason": "Different model/runtime failures with no shared concrete root cause." - }, - { - "left": "issue:44464", - "right": "issue:44977", - "accept": false, - "reason": "Both are generation issues, but one is chunked compiled forward and the other is flash-attention behavior." - }, - { - "left": "issue:28282", - "right": "issue:35141", - "accept": false, - "reason": "ImportError for missing PyTorch vs embedding reinitialization after resize are unrelated." - }, - { - "left": "issue:43643", - "right": "issue:44561", - "accept": false, - "reason": "Both mention trust_remote_code, but one is missing config fields and the other is removal of a utility function." - }, - { - "left": "issue:45491", - "right": "issue:46002", - "accept": false, - "reason": "Both mention NaNs, but they affect different models and different parts of the stack." - }, { "left": "issue:43504", "right": "issue:43720", "accept": false, - "reason": "Different loading bugs for different model families and different failure modes." + "reason": "Different bugs: legacy field loading for Beit vs BitNet packed-weight unpacking during accelerate loading." }, { "left": "issue:43504", "right": "issue:43550", "accept": false, - "reason": "Legacy-field load failure is not the same as torch.compile/SDPA failure." + "reason": "Different models and failures: Beit pretrained loading vs Bamba torch.compile/SDPA behavior." }, { "left": "issue:43504", "right": "issue:44610", "accept": false, - "reason": "Model preset loading bug vs processor/model size mismatch." + "reason": "One is a legacy-field load error; the other is an OmDet processor size mismatch." }, { "left": "issue:44466", "right": "issue:46050", "accept": false, - "reason": "Both involve save/load behavior, but the reported serialization failures are different." + "reason": "Both involve serialization, but one is tied weights/device-dependent save behavior and the other is sinq quantized save/load breakage." }, { "left": "issue:33666", "right": "issue:36010", "accept": false, - "reason": "Multi-GPU training issue vs import path breakage are unrelated." + "reason": "Multi-GPU Qwen2-VL training issues are unrelated to the GenerationMixin import regression." }, { "left": "issue:43742", "right": "issue:44821", "accept": false, - "reason": "KeyError while loading a model vs inability to load an image processor from URL are different bugs." + "reason": "A model key error and loading an AutoImageProcessor from URL are different load paths and failures." }, { "left": "issue:32090", "right": "issue:33666", "accept": false, - "reason": "Trainer broadcast TypeError is unrelated to Qwen2-VL multi-GPU training." + "reason": "Trainer GPU broadcast NoneType error is a different distributed-training failure from Qwen2-VL multi-GPU training." }, { "left": "issue:44977", "right": "issue:46209", "accept": false, - "reason": "Different inference failures: flash-attention generation vs FP8 on B200." + "reason": "Different runtime paths: Qwen3.5 flash-attention generation vs FP8 inference on B200." }, { "left": "issue:43577", "right": "issue:45588", "accept": false, - "reason": "Dtype propagation in BLIP2 is unrelated to flash_attention.py crashing on sink-less models." + "reason": "Blip2 dtype propagation and flash_attention s_aux=None crash are unrelated code paths." }, { "left": "issue:43688", "right": "issue:45440", "accept": false, - "reason": "Different MoE/model-divergence bugs, not the same underlying defect." + "reason": "Aux-loss normalization in OLMoE/GPT Oss is not the same bug as DeepSeekV3MoE divergence." }, { "left": "issue:41720", "right": "issue:44464", "accept": false, - "reason": "CUDA assert during device-mapped Qwen3 load is not the same as compiled-forward chunked generation inconsistency." + "reason": "Auto device mapping cudaErrorAssert and chunked-generation compiled-forward inconsistency are distinct failures." }, { "left": "issue:28282", "right": "issue:32090", "accept": false, - "reason": "Missing PyTorch ImportError vs Trainer broadcast NoneType TypeError are unrelated." + "reason": "PyTorch missing ImportError and Trainer _gpu_broadcast_one TypeError are unrelated setup vs distributed-runtime problems." }, { "left": "issue:43720", "right": "issue:43927", "accept": false, - "reason": "Accelerate unpacking bug for BitNet is unrelated to config token-ID loss." + "reason": "BitNet packed-weight loading and DiaConfig token-ID persistence are unrelated bugs." }, { "left": "issue:43262", "right": "issue:43295", "accept": false, - "reason": "Both involve processors, but one is sampling-rate defaulting and the other is tokenizer/image passthrough regression." + "reason": "Audio chat-template sampling-rate default and processor.tokenizer/image regression are different processor behaviors." }, { "left": "issue:43606", "right": "issue:43927", "accept": false, - "reason": "CPU offload device mismatch is unrelated to save/load token-ID persistence." + "reason": "CPU offload device mismatch and custom token ID loss are unrelated issues." }, { "left": "issue:43452", "right": "issue:44451", "accept": false, - "reason": "Both are loading issues, but they concern different loaders and different failure modes." + "reason": "General gguf/from_pretrained loading breakage is not the same as ScandiBERT-specific load failure." }, { "left": "issue:45823", "right": "issue:46050", "accept": false, - "reason": "Device-mapping mismatch and quantized save/load breakage are different bugs." + "reason": "Gemma4 auto device-map device mismatch is unrelated to sinq quantization save/load corruption." }, { "left": "issue:33666", "right": "issue:36331", "accept": false, - "reason": "Multi-GPU training issue vs CustomTrainer signature mismatch are unrelated." + "reason": "Qwen2-VL multi-GPU training and custom trainer compute_loss signature mismatch are different regressions." }, { "left": "issue:43994", "right": "issue:46192", "accept": false, - "reason": "Nonsensical SigLIP2 outputs vs ignored return_tensors are different bugs." + "reason": "SigLIP2 wrong outputs and return_tensors being ignored are different API bugs." }, { "left": "issue:43525", "right": "issue:43572", "accept": false, - "reason": "Both are missing config fields, but they are different fields in different models." + "reason": "Similar theme of missing config fields, but different models and different missing attributes." }, { "left": "issue:43720", "right": "issue:44361", "accept": false, - "reason": "Different model-loading failures with no shared root cause." + "reason": "BitNet packed-weight loading is unrelated to MLukeTokenizer AttributeError on tasks." }, { "left": "issue:44849", "right": "issue:46209", "accept": false, - "reason": "Qwen3.5 hidden-state bug is unrelated to FP8 inference failure." + "reason": "Qwen3.5 hidden-state output bug and FP8 B200 inference failure are unrelated." }, { "left": "issue:43493", "right": "issue:43825", "accept": false, - "reason": "SigLIP2 implementation mismatch and pipeline translation-message issue are unrelated." + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message wording are not the same bug." }, { "left": "issue:44964", "right": "issue:45405", "accept": false, - "reason": "Model loading failure vs dependency version pinning are not duplicates." + "reason": "Phi-4 multimodal load failure and MIN_PEFT_VERSION packaging issue are unrelated." }, { "left": "issue:39692", "right": "issue:43493", "accept": false, - "reason": "Docs/example errors are unrelated to a core implementation discrepancy." + "reason": "Docs-example errors and core SigLIP2 implementation discrepancy are different issues." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video-input regression across Qwen models is separate from Qwen3.5 num_labels propagation." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Blip2 dtype loading and MPS attention correctness differ in subsystem and symptom." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Hidden-states output bug and incorrect visual-encoder save_pretrained keys are distinct Qwen3.5 regressions." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "Different tokenizer families and failures; mdeberta special tokens vs DeepSeek-Coder v1 output regression." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Gemma4 text-only token-type defaults and Gemma4 31B multi-GPU OOM are unrelated problems." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "SigLIP2 documentation errors and processor.tokenizer/image regression do not share the same code-path bug." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are different training/loading issues." } ] }, @@ -7554,10 +7532,8 @@ "nodes": [ "issue:33357", "issue:33666", - "issue:39692", "issue:43257", "issue:43278", - "issue:43295", "issue:43299", "issue:43334", "issue:43493", @@ -7565,50 +7541,44 @@ "issue:43540", "issue:43572", "issue:43575", - "issue:43577", "issue:43606", "issue:43646", + "issue:43653", "issue:43688", - "issue:43749", "issue:43824", "issue:43844", "issue:43927", + "issue:44112", "issue:44291", "issue:44361", + "issue:44423", "issue:44464", "issue:44466", - "issue:44479", "issue:44488", "issue:44493", "issue:44554", "issue:44568", "issue:44589", "issue:44625", + "issue:44734", "issue:44792", "issue:44843", - "issue:44849", "issue:44977", - "issue:45200", "issue:45229", - "issue:45357", "issue:45405", "issue:45520", + "issue:45588", + "issue:45701", "issue:45704", "issue:45759", "issue:45812", "issue:45823", + "issue:46050", "issue:46285", - "issue:46424", - "issue:46489" + "issue:46365", + "issue:46424" ], - "soft_pairs": [ - "issue:44479|issue:44625", - "issue:43577|issue:44554", - "issue:44849|issue:45357", - "issue:44568|issue:46489", - "issue:45200|issue:45229", - "issue:39692|issue:43295", - "issue:43688|issue:43749", + "soft_pairs": [ "issue:44488|issue:44843", "issue:43540|issue:44977", "issue:44568|issue:45812", @@ -7629,204 +7599,211 @@ "issue:44792|issue:45405", "issue:44493|issue:46285", "issue:43257|issue:43299", - "issue:43575|issue:45405" + "issue:43575|issue:45405", + "issue:43504|issue:43653", + "issue:43606|issue:44112", + "issue:44625|issue:45759", + "issue:44554|issue:46365", + "issue:43688|issue:45701", + "issue:45588|issue:46050", + "issue:44423|issue:44734" ], - "prepared_review_unit_hash": "98a1dd4157c80931424211c49e79269df3238305522c1433a1d7435b148f782b", + "prepared_review_unit_hash": "a9bba81ac26be61bd4560fb52d8474e5b1a38b42a6683c09078ea55b4988dfbb", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 28, - "serialized_chars": 36288, - "estimated_input_tokens": 9072, - "estimated_eval_tokens": 18400 + "serialized_chars": 35685, + "estimated_input_tokens": 8922, + "estimated_eval_tokens": 18100 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a loose bundle of mostly unrelated Transformers bug reports, with several small tokenizer/config/model-specific similarity pairs but no pair that clearly describes the same underlying bug. I would not merge any of the soft candidates; the cluster is better treated as multiple separate subtopics.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 45812 is the strongest representative of the cluster because it is a broad, high-impact v5 tokenizer regression with multiple inbound references, but the overall set is too heterogeneous to have a true single duplicate canonical.", + "summary": "This cluster is very heterogeneous: most items are isolated Transformers bugs across different models, configs, and failure modes. The soft-similarity links are all false positives; none look like the same underlying bug or fixable in one PR.", + "confidence": 0.97, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "Issue 45812 is the best single issue to anchor this cluster: it is concrete, broadly applicable, and closer to a cross-cutting regression than the more model-specific reports.", + "best_issue_reason": "No single issue cleanly represents the cluster because the items span unrelated bugs; if forced, the closest broad umbrella is still not a true canonical duplicate target.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44479", - "right": "issue:44625", - "accept": false, - "reason": "Both are Qwen3.5-related, but one is a video-input regression and the other is a num_labels propagation bug; different code paths and fixes." - }, - { - "left": "issue:43577", - "right": "issue:44554", - "accept": false, - "reason": "One is a dtype-loading bug in BLIP2, the other is an MPS attention correctness issue; no shared underlying defect." - }, - { - "left": "issue:44849", - "right": "issue:45357", - "accept": false, - "reason": "Both involve Qwen3.5, but one is an output_hidden_states bug and the other is incorrect save_pretrained visual encoder keys; separate problems." - }, - { - "left": "issue:44568", - "right": "issue:46489", - "accept": false, - "reason": "Both are tokenizer regressions, but they affect different models and behaviors; not the same concrete bug." - }, - { - "left": "issue:45200", - "right": "issue:45229", - "accept": false, - "reason": "Gemma4 fine-tuning token-type defaults vs multi-GPU inference OOM are unrelated issues." - }, - { - "left": "issue:39692", - "right": "issue:43295", - "accept": false, - "reason": "SigLIP2 doc/example errors and a processor.tokenizer regression are different failures with different fixes." - }, - { - "left": "issue:43688", - "right": "issue:43749", - "accept": false, - "reason": "Auxiliary-loss normalization and FSDP CPU RAM efficient loading are distinct training/runtime bugs." - }, { "left": "issue:44488", "right": "issue:44843", "accept": false, - "reason": "One is a model load failure for a specific checkpoint, the other is an offline-mode regression in tokenizer regex patching; different root causes." + "reason": "Both are loading-related, but one is a specific model load failure and the other is an offline-mode hub call regression; different code paths and fixes." }, { "left": "issue:43540", "right": "issue:44977", "accept": false, - "reason": "Both are Qwen3.5-family, but one is video processing and the other is flash-attention generation behavior; not the same bug." + "reason": "Different Qwen3 variants and symptoms: one is a video-input ValueError in Qwen3OmniMoe, the other is flash-attention generation quality in Qwen3.5." }, { "left": "issue:44568", "right": "issue:45812", "accept": false, - "reason": "Both are tokenizer regressions, but they hit different models and symptoms; one is special-token handling, the other incorrect token IDs for Granite." + "reason": "Both are tokenizer regressions, but they affect different model families and different tokenization behavior." }, { "left": "issue:44464", "right": "issue:46424", "accept": false, - "reason": "Chunked generation inconsistency with compiled forward is unrelated to cache_config max_cache_length behavior." + "reason": "Chunked generation compiled-forward inconsistency is unrelated to cache max length handling for quantized cache." }, { "left": "issue:43493", "right": "issue:43844", "accept": false, - "reason": "SigLIP2 HF/JAX discrepancy and ZeRO-3 gradient explosion are unrelated correctness issues." + "reason": "One is a model-implementation discrepancy vs JAX; the other is a DeepSpeed training instability issue." }, { "left": "issue:43646", "right": "issue:44466", "accept": false, - "reason": "Custom model initialization breakage and device-dependent lm_head serialization are different v5 regressions." + "reason": "Generic custom model initialization breakage is not the same bug as device-dependent lm_head serialization inconsistency." }, { "left": "issue:43334", "right": "issue:43572", "accept": false, - "reason": "Qwen3-VL pad_token_id missing and StableLmConfig pad_token_idx missing are similar class of config issues but clearly different model/config paths." + "reason": "Both mention missing config fields, but they are different fields on different model classes with distinct failure modes." }, { "left": "issue:43824", "right": "issue:44589", "accept": false, - "reason": "Import failure for a specific Qwen2.5-VL class and Float8 storage lookup error are unrelated import/runtime bugs." + "reason": "Importing a missing VL class and failing on Float8 storage lookup are unrelated failures." }, { "left": "issue:33357", "right": "issue:33666", "accept": false, - "reason": "MacOS bus error with CLIP and Qwen2-VL multi-GPU training are entirely different problems." + "reason": "MacOS CLIP bus error and Qwen2-VL multi-GPU training are entirely different problems." }, { "left": "issue:44589", "right": "issue:45520", "accept": false, - "reason": "Float8 storage type lookup and flash_attn import_utils KeyError are separate backend/import issues." + "reason": "Float8 storage lookup error is unrelated to a Python 3.13 flash_attn import-utils KeyError." }, { "left": "issue:43688", "right": "issue:45704", "accept": false, - "reason": "Aux-loss normalization and T5 Apex FusedRMSNorm memory leak are unrelated model-specific bugs." + "reason": "Aux-loss normalization in MoE models is unrelated to a T5 apex RMSNorm memory leak." }, { "left": "issue:43334", "right": "issue:45759", "accept": false, - "reason": "Missing pad_token_id in Qwen3-VL config and AutoModelForCausalLM.from_config not unwrapping text_config are different composite-config issues." + "reason": "Qwen3-VL pad_token_id loss and composite Qwen 3.5/3.6 text_config unwrapping are different configuration bugs." }, { "left": "issue:43278", "right": "issue:43844", "accept": false, - "reason": "Embedding dtype mismatch in eval and ZeRO-3 gradient growth are not the same defect." + "reason": "Embedding dtype drift during evaluate is unrelated to the ZeRO-3 gradient increase issue." }, { "left": "issue:43504", "right": "issue:43606", "accept": false, - "reason": "Legacy field blocking BEiT preset load and Bark CPU-offload device mismatch are unrelated load/runtime failures." + "reason": "Legacy-field model loading failure and CPU-offload device mismatch do not share a concrete code path." }, { "left": "issue:43927", "right": "issue:44361", "accept": false, - "reason": "DiaConfig token ID persistence and MLukeTokenizer task AttributeError are separate tokenizer/config bugs." + "reason": "DiaConfig token ID persistence and MLukeTokenizer task AttributeError are different model/tokenizer issues." }, { "left": "issue:45229", "right": "issue:45823", "accept": false, - "reason": "Gemma4 multi-GPU OOM and Gemma4 PLE device mismatch share a model family but not the same failure mode." + "reason": "Both are Gemma4, but one is OOM under multi-GPU inference while the other is a forward-time device mismatch." }, { "left": "issue:43572", "right": "issue:44291", "accept": false, - "reason": "StableLm pad_token_idx regression and unexpected _is_hf_initialized argument in init_empty_weights are distinct config/API breakages." + "reason": "Both are v5 compatibility issues, but they concern different APIs: missing pad_token_idx vs unexpected _is_hf_initialized in init_empty_weights." }, { "left": "issue:44792", "right": "issue:45405", "accept": false, - "reason": "Janus test failure and MIN_PEFT_VERSION bump are not the same underlying bug." + "reason": "A failing Janus image-generation test is unrelated to a PEFT version pin bump." }, { "left": "issue:44493", "right": "issue:46285", "accept": false, - "reason": "Unexpected position-id keys and CLIP text model structural inconsistency are different model-structure regressions." + "reason": "Unexpected position-id keys in many models is not the same as the CLIP text-model structure change." }, { "left": "issue:43257", "right": "issue:43299", "accept": false, - "reason": "Qwen3 MoE weights not converted with accelerate+deepspeed and Qwen3VL MoE loading breakage are related by family only, not the same concrete issue." + "reason": "Qwen3 MoE weight conversion with accelerate/deepspeed and Qwen3VL MoE load failure are related only by family name, not the same bug." }, { "left": "issue:43575", "right": "issue:45405", "accept": false, - "reason": "TP loading OOM for Qwen2-57B and a PEFT version bump are unrelated." + "reason": "Tensor-parallel OOM on a Qwen2 model is unrelated to an unreleased dependency-version bump." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Beit pretrained load failure and BigBirdTokenizer special-token registration are different subsystems and bugs." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "CPU offload device mismatch and a stale GraniteSpeech CI test are not the same underlying issue." + }, + { + "left": "issue:44625", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and composite Qwen config unwrapping are distinct config-serialization problems." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "An MPS attention correctness bug is unrelated to Trainer import failure caused by datasets installation." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "MoE auxiliary-loss normalization and a broad tokenization regression report are not the same concrete bug." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention sink-less-model crash and sinq save/load quantization breakage are unrelated." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both involve serve, but one crashes on multimodal continuous batching and the other on KV-cache continuation indexing." } ] }, @@ -7856,57 +7833,50 @@ "issue:29942", "issue:30333", "issue:32090", + "issue:33357", "issue:34567", "issue:38175", + "issue:39692", "issue:43065", "issue:43232", "issue:43278", + "issue:43295", + "issue:43450", "issue:43493", - "issue:43504", "issue:43540", - "issue:43606", - "issue:43653", + "issue:43550", + "issue:43572", "issue:43688", "issue:43701", "issue:43844", "issue:43873", "issue:43994", - "issue:44112", "issue:44186", - "issue:44423", + "issue:44466", "issue:44493", "issue:44521", "issue:44554", - "issue:44625", - "issue:44734", + "issue:44610", "issue:44849", "issue:44964", + "issue:44987", "issue:45072", "issue:45084", + "issue:45127", "issue:45198", "issue:45200", "issue:45290", + "issue:45356", "issue:45406", "issue:45446", "issue:45520", - "issue:45588", - "issue:45701", - "issue:45759", "issue:45823", "issue:46050", "issue:46121", "issue:46192", - "issue:46365", "issue:46369" ], "soft_pairs": [ - "issue:43504|issue:43653", - "issue:43606|issue:44112", - "issue:44625|issue:45759", - "issue:44554|issue:46365", - "issue:43688|issue:45701", - "issue:45588|issue:46050", - "issue:44423|issue:44734", "issue:29942|issue:30333", "issue:43844|issue:44964", "issue:43994|issue:44493", @@ -7927,204 +7897,213 @@ "issue:43232|issue:44849", "issue:43493|issue:43701", "issue:45406|issue:46121", - "issue:45823|issue:46192" + "issue:45823|issue:46192", + "issue:43295|issue:43450", + "issue:28218|issue:33357", + "issue:43295|issue:44610", + "issue:44554|issue:45356", + "issue:43295|issue:43873", + "issue:43572|issue:44987", + "issue:39692|issue:46192", + "issue:43550|issue:43844", + "issue:44466|issue:45127" ], - "prepared_review_unit_hash": "8c543ea7c28d459463ae4ac6e16d47a49b354693b18bc68f11d46e5ef3edb72d", + "prepared_review_unit_hash": "708fa07163da345887dd3c38399925f8b4976c4ec60fa3ecd489af3102e711f2", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 36100, - "estimated_input_tokens": 9025, - "estimated_eval_tokens": 18306 + "soft_pair_count": 30, + "serialized_chars": 36687, + "estimated_input_tokens": 9172, + "estimated_eval_tokens": 18600 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a loose collection of unrelated bug reports. None of the proposed soft pairs look like the same underlying issue/change, so all soft edges should be rejected.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "No soft edge pair looks like a true duplicate; the set is a grab-bag of unrelated tokenizer, processor, quantization, generation, and training issues rather than one coherent bug cluster.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue: the items are heterogeneous and the soft-pair similarities are superficial.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "No issue is a convincing cluster representative because none of the candidates share the same underlying defect.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43504", - "right": "issue:43653", - "accept": false, - "reason": "Different failure modes: legacy pretrained-field loading for BEiT vs BigBirdTokenizer special-token registration/empty decode." - }, - { - "left": "issue:43606", - "right": "issue:44112", - "accept": false, - "reason": "Unrelated bugs: CPU offload device mismatch in suno/bark-small vs a stale GraniteSpeech CI test override." - }, - { - "left": "issue:44625", - "right": "issue:45759", - "accept": false, - "reason": "Qwen config propagation of num_labels is a different code path from unwrapping text_config in AutoModelForCausalLM.from_config." - }, - { - "left": "issue:44554", - "right": "issue:46365", - "accept": false, - "reason": "MPS attention correctness issue is unrelated to importing Trainer failing because of the datasets package." - }, - { - "left": "issue:43688", - "right": "issue:45701", - "accept": false, - "reason": "Auxiliary-loss normalization bug is not the same as a version-dependent tokenization regression." - }, - { - "left": "issue:45588", - "right": "issue:46050", - "accept": false, - "reason": "flash_attention.py crashing on s_aux=None is unrelated to sinq save/load quantization breakage." - }, - { - "left": "issue:44423", - "right": "issue:44734", - "accept": false, - "reason": "Both involve serve, but one is a multimodal continuous-batching crash and the other is KV-cache continuation indexing." - }, { "left": "issue:29942", "right": "issue:30333", "accept": false, - "reason": "Flash Attention 2 test failures and MLFlow reporting on failed jobs are different subsystems and bugs." + "reason": "Different failures in different areas: FlashAttention test regressions vs MLflow reporting on failed jobs." }, { "left": "issue:43844", "right": "issue:44964", "accept": false, - "reason": "ZeRO-3 gradient explosion during training is not the same as a model-loading failure for Phi-4 multimodal." + "reason": "Unrelated bugs: ZeRO-3 gradient growth vs model loading failure for Phi-4 multimodal." }, { "left": "issue:43994", "right": "issue:44493", "accept": false, - "reason": "SigLIP2 nonsensical outputs and unexpected position-id keys point to different model bugs." + "reason": "SigLIP2 output quality issue vs position-id key mismatch; different code paths and symptoms." }, { "left": "issue:43688", "right": "issue:43873", "accept": false, - "reason": "Aux-loss normalization is unrelated to offloading behavior with quantization." + "reason": "Aux-loss normalization bug is unrelated to quantization/offloading behavior." }, { "left": "issue:29159", "right": "issue:32090", "accept": false, - "reason": "Slow/fast tokenizer inconsistency is a different issue from NoneType being passed to _gpu_broadcast_one." + "reason": "Tokenizer slow/fast inconsistency vs Trainer GPU broadcast TypeError are unrelated." }, { "left": "issue:28218", "right": "issue:34567", "accept": false, - "reason": "Extra whitespace after added tokens is unrelated to TrainerState.num_input_tokens_seen not updating." + "reason": "Added-token spacing bug is unrelated to TrainerState token counting." }, { "left": "issue:45084", "right": "issue:45520", "accept": false, - "reason": "Non-template node compilation failure is not the same as a Python 3.13 flash_attn import KeyError." + "reason": "Template-node compilation error and flash_attn import KeyError are separate issues." }, { "left": "issue:44554", "right": "issue:45446", "accept": false, - "reason": "MPS attention correctness bug is unrelated to an incorrect PyTorch version check in flex_attention." + "reason": "Both mention attention-related code, but one is an MPS correctness bug and the other is a PyTorch version guard/import check." }, { "left": "issue:15354", "right": "issue:38175", "accept": false, - "reason": "torch.jit.script GeneratorExp export failure is unrelated to unexpected zero probabilities in SigLIP2." + "reason": "TorchScript GeneratorExp export failure vs SigLIP2 zero probabilities are unrelated." }, { "left": "issue:43540", "right": "issue:45200", "accept": false, - "reason": "Qwen3OmniMoe video-input processing error is unrelated to Gemma4 mm_token_type_ids defaulting." + "reason": "Video input processing error in Qwen3OmniMoe vs Gemma4 multimodal token-type defaulting." }, { "left": "issue:43065", "right": "issue:46369", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is not the same as a memory overflow in sam3-litetext." + "reason": "Dummy Conv2d in SAM3 pixel decoder is not the same as memory overflow in sam3-litetext." }, { "left": "issue:43278", "right": "issue:43701", "accept": false, - "reason": "Embedding dtype drift between train/eval is unrelated to resume_from_checkpoint key mismatches." + "reason": "BF16/FP32 dtype drift during evaluate is unrelated to resume_from_checkpoint key mismatch." }, { "left": "issue:43873", "right": "issue:46050", "accept": false, - "reason": "Quantization/offloading not working as expected is different from a broken sinq save_pretrained/from_pretrained flow." + "reason": "Quantization/offloading symptoms differ: offloading not working vs broken save/load round-trip after sinq quantization." }, { "left": "issue:45072", "right": "issue:45198", "accept": false, - "reason": "bfloat16 dtype mismatches in inference are not the same as Wav2Vec2 save_pretrained/tokenization failures." - }, - { - "left": "issue:45200", - "right": "issue:45823", - "accept": false, - "reason": "Text-only Gemma4 mm_token_type_ids defaults are unrelated to a device_map='auto' forward device mismatch." - }, - { - "left": "issue:44186", - "right": "issue:46192", - "accept": false, - "reason": "LayoutLMv2 tokenizer crashes on NER/padding are a different bug from return_tensors being ignored when text_kwargs is passed." + "reason": "bfloat16 dtype mismatches in inference are unrelated to Wav2Vec2 save/tokenization failures." }, { "left": "issue:44521", "right": "issue:45290", "accept": false, - "reason": "All-zero assistant masks for multimodal inputs are not the same as crashing on tool-call assistant messages with no content." + "reason": "Both touch apply_chat_template, but one is zero assistant masks for multimodal inputs and the other is a crash on tool-call assistant messages; different edge cases." }, { "left": "issue:43232", "right": "issue:44849", "accept": false, - "reason": "Generation kwargs after sync_gpus and Qwen3.5 output_hidden_states behavior are separate generation bugs." + "reason": "Generation kwargs update after sync_gpus is unrelated to Qwen3.5 output_hidden_states bug." }, { "left": "issue:43493", "right": "issue:43701", "accept": false, - "reason": "SigLIP2 HF-vs-JAX discrepancy is unrelated to a resume_from_checkpoint key mismatch." + "reason": "SigLIP2 HF-vs-JAX discrepancy is unrelated to checkpoint resume key mismatch." }, { "left": "issue:45406", "right": "issue:46121", "accept": false, - "reason": "Gemma4Processor missing _tokenizer in serve is unrelated to convert_rope_params_to_dict failing on a JSON-loaded list." + "reason": "Gemma4Processor _tokenizer crash and rope-params list TypeError are different bugs." }, { "left": "issue:45823", "right": "issue:46192", "accept": false, - "reason": "Gemma4 PLE device mismatch and return_tensors being ignored affect different code paths and outcomes." + "reason": "Device-map forward mismatch is unrelated to return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression is unrelated to batched video processor shape errors." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Added-token spacing bug is unrelated to MacOS bus error in CLIP model loading." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Custom processor/tokenizer regression is not the same as OmDet-Turbo\u2019s 640px-vs-224px preprocessing mismatch." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "Different attention/tokenizer problems: MPS attention correctness vs Kimi-K2.5 tokenizer regression." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression is unrelated to quantization offloading behavior." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is not the same as loading failure for physical-intelligence/fast." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 doc example errors are unrelated to return_tensors being ignored with text_kwargs." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "torch.compile+SDPA failure for Bamba-9B-v2 is unrelated to ZeRO-3 gradient blow-up." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Both involve weights, but one is inconsistent serialization and the other is LoRA merge collapse with extended vocab." } ] }, @@ -8148,11 +8127,8 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:28218", - "issue:33357", "issue:37428", "issue:38175", - "issue:39692", "issue:42490", "issue:42915", "issue:43262", @@ -8163,50 +8139,44 @@ "issue:43493", "issue:43525", "issue:43526", - "issue:43550", "issue:43572", "issue:43575", "issue:43653", "issue:43696", + "issue:43819", "issue:43824", - "issue:43844", + "issue:43827", "issue:43856", - "issue:43873", + "issue:43901", "issue:43950", "issue:43957", "issue:43975", "issue:44112", "issue:44442", - "issue:44466", - "issue:44554", + "issue:44514", + "issue:44521", + "issue:44561", "issue:44610", "issue:44857", "issue:44898", - "issue:44987", - "issue:45127", + "issue:45083", "issue:45198", "issue:45200", + "issue:45290", "issue:45325", - "issue:45356", "issue:45381", "issue:45412", "issue:45520", + "issue:45759", "issue:46170", "issue:46178", "issue:46192", "issue:46209", - "issue:46285" + "issue:46234", + "issue:46285", + "issue:46326" ], "soft_pairs": [ - "issue:43295|issue:43450", - "issue:28218|issue:33357", - "issue:43295|issue:44610", - "issue:44554|issue:45356", - "issue:43295|issue:43873", - "issue:43572|issue:44987", - "issue:39692|issue:46192", - "issue:43550|issue:43844", - "issue:44466|issue:45127", "issue:43572|issue:43957", "issue:44898|issue:45412", "issue:43950|issue:46285", @@ -8229,228 +8199,223 @@ "issue:38175|issue:45412", "issue:45520|issue:46170", "issue:43262|issue:43450", - "issue:44610|issue:44857" + "issue:44610|issue:44857", + "issue:44561|issue:46234", + "issue:45759|issue:46285", + "issue:43819|issue:44112", + "issue:43827|issue:43901", + "issue:45290|issue:46326", + "issue:44514|issue:44521", + "issue:43819|issue:45083" ], - "prepared_review_unit_hash": "fa97e3e389ca713f95dd870e7e79780b934001f5e61c2b2fbae54defe867283a", + "prepared_review_unit_hash": "7b505cebe9dc030015e3494c0820482268fd2e3afc67eb2964d41f835ef72f86", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37614, - "estimated_input_tokens": 9404, - "estimated_eval_tokens": 19064 + "soft_pair_count": 30, + "serialized_chars": 36973, + "estimated_input_tokens": 9244, + "estimated_eval_tokens": 18744 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose grab-bag of mostly unrelated Transformers bug reports. A few pairs share broad vocabulary (tokenizers, processors, multimodal models), but the titles point to different concrete failure modes, so I would not merge any of the soft edges.", - "confidence": 0.87, - "canonical_issue_reason": "issue:43295 is the best anchor because it is a clear regression report and one of the most connected items in the soft-similarity graph, but it is still distinct from the other reports.", + "summary": "This is not a single duplicate cluster; it\u2019s a grab bag of mostly unrelated Transformers regressions, with only a loose multimodal/chat-template subtheme. None of the soft pairs are close enough to merge as true duplicates, so I\u2019d treat them as distinct issues and use issue 44521 only as the broadest representative of the chat-template subgroup.", + "confidence": 0.89, + "canonical_issue_reason": "If a representative issue is needed, 44521 is the broadest and most active multimodal apply_chat_template regression in this set, but the cluster overall is not a true duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43295 is the strongest standalone issue in this set: it describes a specific user-facing regression, is easy to understand, and is the closest thing here to a central duplicate target.", + "best_issue_reason": "44521 is the most central issue among the few related items: it has the broadest multimodal chat-template scope, active discussion, and references, making it the best stand-in for the loose subgroup.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43295", - "right": "issue:43450", - "accept": false, - "reason": "Processor/tokenizer regression vs batched video-shape bug; different code paths and symptoms." - }, - { - "left": "issue:28218", - "right": "issue:33357", - "accept": false, - "reason": "Tokenizer spacing bug vs MacOS bus error in a CLIP model; unrelated failures." - }, - { - "left": "issue:43295", - "right": "issue:44610", - "accept": false, - "reason": "Custom processor/tokenizer regression vs OmDet-Turbo input-size mismatch; not the same bug." - }, - { - "left": "issue:44554", - "right": "issue:45356", - "accept": false, - "reason": "MPS attention correctness issue vs tokenizer codec/warning regression; different subsystems and failure modes." - }, - { - "left": "issue:43295", - "right": "issue:43873", - "accept": false, - "reason": "Processor/tokenizer API regression vs quantization offloading issue; too far apart to be one fix." - }, - { - "left": "issue:43572", - "right": "issue:44987", - "accept": false, - "reason": "Missing pad_token_idx in StableLmConfig vs loading a specific model failing; same broad area, different concrete problem." - }, - { - "left": "issue:39692", - "right": "issue:46192", - "accept": false, - "reason": "Documentation/example errors vs return_tensors handling bug; unrelated." - }, - { - "left": "issue:43550", - "right": "issue:43844", - "accept": false, - "reason": "torch.compile + SDPA failure in one model vs gradient explosion with ZeRO-3; different bugs." - }, - { - "left": "issue:44466", - "right": "issue:45127", - "accept": false, - "reason": "Serialization of tied weights vs collapse after LoRA merge with extended vocab; not the same code-path." - }, { "left": "issue:43572", "right": "issue:43957", "accept": false, - "reason": "Config attribute regression for StableLm vs meta-device loading breakage; both loading-related but distinct." + "reason": "Both are v5 loading regressions, but StableLm missing pad_token_idx and meta-device loading are different root causes." }, { "left": "issue:44898", "right": "issue:45412", "accept": false, - "reason": "Perceiver interpolation failure vs RT-DETR memory leak; no shared underlying bug." + "reason": "Different bugs: Perceiver non-default-resolution interpolation vs RT-DETR memory not being released." }, { "left": "issue:43950", "right": "issue:46285", "accept": false, - "reason": "Non-persistent buffer corruption during from_pretrained vs CLIP class structure inconsistency; different regressions." + "reason": "Both involve model-loading structure changes, but buffer persistence corruption and CLIP text-model layout inconsistency are not the same defect." }, { "left": "issue:45520", "right": "issue:46178", "accept": false, - "reason": "flash_attn import KeyError on Python 3.13 vs tensorflow_text backend mapping error; unrelated import issues." + "reason": "Separate import/backends failures: flash_attn lookup on Python 3.13 vs tensorflow_text not in BACKENDS_MAPPING." }, { "left": "issue:45381", "right": "issue:46209", "accept": false, - "reason": "Qwen2.5-VL vision position IDs bug vs FP8 inference failure on B200; different models and failure modes." + "reason": "Unrelated code paths: Qwen2.5-VL rope/position ids vs FP8 inference failure on B200." }, { "left": "issue:45200", "right": "issue:45325", "accept": false, - "reason": "Gemma 4 multimodal token type defaults vs Qwen2.5-VL rope scaling bug; not the same issue." + "reason": "Both touch multimodal defaults, but Gemma 4 mm_token_type_ids and Qwen2.5-VL temporal position ids are distinct issues." }, { "left": "issue:43450", "right": "issue:43575", "accept": false, - "reason": "Video processor batch-shape bug vs large-model OOM under tensor parallelism; unrelated." + "reason": "Video processor batch-shape bug and tensor-parallel OOM are unrelated." }, { "left": "issue:37428", "right": "issue:43957", "accept": false, - "reason": "Missing flash-support import vs meta-device loading regression; both loading-related but not the same bug." + "reason": "ImportError for a flash-attention helper is not the same as meta-device model loading breakage." }, { "left": "issue:43334", "right": "issue:43525", "accept": false, - "reason": "Qwen3-VL missing pad_token_id vs Llama4 missing pad_token_id; similar symptom, but different model/config paths." + "reason": "Same symptom name, different models/config classes; no evidence of one underlying bug." }, { "left": "issue:43824", "right": "issue:43975", "accept": false, - "reason": "ImportError for Qwen2_5_VL class vs DeepSeek coder detokenization bug; completely different issues." + "reason": "Import error for Qwen2.5-VL class vs DeepSeek detokenization regression are different problems." }, { "left": "issue:38175", "right": "issue:43493", "accept": false, - "reason": "SigLIP2 zero probabilities vs broader HF-vs-JAX discrepancy; related model family, but not clearly the same concrete bug." + "reason": "Both are SigLIP2-related, but one is zero probabilities and the other is a broader HF-vs-JAX discrepancy; not enough to treat as one concrete bug." }, { "left": "issue:43653", "right": "issue:44112", "accept": false, - "reason": "BigBirdTokenizer special-token registration bug vs GraniteSpeech stale CI test; different artifacts." + "reason": "Tokenizer special-token registration bug vs GraniteSpeech CI test flake are unrelated." }, { "left": "issue:42490", "right": "issue:43696", "accept": false, - "reason": "Trailing-slash local path import warning vs CUDA OOM on GPT-oss; unrelated." + "reason": "Local-path import warning and GPT-oss OOM are unrelated failure modes." }, { "left": "issue:43278", "right": "issue:43856", "accept": false, - "reason": "BF16-to-FP32 dtype change in evaluate vs Qwen3 MoE memory inefficiency; different regressions." + "reason": "BF16-to-FP32 dtype drift during eval is not the same as MoE training memory inefficiency." }, { "left": "issue:43526", "right": "issue:46192", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug vs return_tensors being ignored; unrelated." + "reason": "Both are processor argument-handling bugs, but reduce_labels truncation and return_tensors/text_kwargs precedence are different defects." }, { "left": "issue:42915", "right": "issue:43824", "accept": false, - "reason": "Qwen3Moe FP8 config failure vs Qwen2_5_VL import error; different models and failure modes." + "reason": "Qwen3MoE FP8 config failure and Qwen2.5-VL import error are unrelated." }, { "left": "issue:43295", "right": "issue:43493", "accept": false, - "reason": "Processor/tokenizer API regression vs SigLIP2 implementation discrepancy; same broad multimodal area, but not the same bug." + "reason": "Custom processor.tokenizer regression and SigLIP2 implementation discrepancy do not share the same code-path." }, { "left": "issue:44442", "right": "issue:45198", "accept": false, - "reason": "AutoTokenizer loading failure for one tokenizer vs Wav2Vec2 save_pretrained/tokenization failure; distinct tokenizer bugs." + "reason": "Both are speech tokenization issues, but they affect different tokenizer families and different save/load paths." }, { "left": "issue:37428", "right": "issue:43824", "accept": false, - "reason": "flash_supports_window_size import error vs missing Qwen2_5_VL export; both import-related, but different symbols and causes." + "reason": "Different import errors in different modules/classes." }, { "left": "issue:38175", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 output anomaly vs RT-DETR memory leak; unrelated." + "reason": "SigLIP2 output anomaly and RT-DETR memory retention are unrelated." }, { "left": "issue:45520", "right": "issue:46170", "accept": false, - "reason": "Python 3.13 flash_attn import issue vs Windows ROCm torch.distributed dependency issue; different platform problems." + "reason": "Both are import-related, but the Windows ROCm torch.distributed issue is a different dependency problem." }, { "left": "issue:43262", "right": "issue:43450", "accept": false, - "reason": "Audio chat-template default sampling rate vs video batch-shape bug; different processors and code paths." + "reason": "Audio chat-template sampling-rate default and video batched-shape bug are different processor regressions." }, { "left": "issue:44610", "right": "issue:44857", "accept": false, - "reason": "Processor image-size mismatch vs AMP/CUDA crash in image loss; unrelated." + "reason": "Different model/task bugs: OmDet processor sizing vs LwDetr AMP crash." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Both are missing helper symbols in import_utils, but they are different functions with different downstream breakages." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Both involve composite-model structure, but Qwen multimodal config unwrapping and CLIP text-model layout inconsistency are not the same bug." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "Unrelated: DAC latent reconstruction mismatch vs GraniteSpeech CI device override." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs regressions, but they cover different pipelines/APIs and are not duplicates." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both concern chat-template edge cases, but tool-call/no-content crashes and mistral_common continue_final_message behavior are distinct." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Same area, but one is a batched padding=False crash and the other is all-zero assistant masks; different concrete failures." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "Completely different areas: DAC from_latents vs qwen3_omni_moe helper output lengths." } ] }, @@ -8474,14 +8439,17 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ + "issue:28218", "issue:29942", "issue:32090", + "issue:35141", "issue:37428", "issue:39692", "issue:41720", "issue:42796", "issue:43278", "issue:43335", + "issue:43388", "issue:43425", "issue:43504", "issue:43526", @@ -8489,18 +8457,14 @@ "issue:43606", "issue:43611", "issue:43653", + "issue:43720", "issue:43819", "issue:43825", - "issue:43827", - "issue:43901", "issue:44112", "issue:44206", "issue:44351", "issue:44448", "issue:44493", - "issue:44514", - "issue:44521", - "issue:44561", "issue:44743", "issue:44792", "issue:44855", @@ -8510,27 +8474,21 @@ "issue:45200", "issue:45245", "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", "issue:45563", "issue:45584", "issue:45588", "issue:45593", - "issue:45759", "issue:45823", "issue:45859", "issue:46209", "issue:46234", - "issue:46285", - "issue:46326", "issue:46402" ], "soft_pairs": [ - "issue:44561|issue:46234", - "issue:45759|issue:46285", - "issue:43819|issue:44112", - "issue:43827|issue:43901", - "issue:45290|issue:46326", - "issue:44514|issue:44521", - "issue:43819|issue:45083", "issue:41720|issue:46209", "issue:44857|issue:45593", "issue:43335|issue:43550", @@ -8554,229 +8512,260 @@ "issue:29942|issue:32090", "issue:39692|issue:43526", "issue:45588|issue:45859", - "issue:43335|issue:45083" + "issue:43335|issue:45083", + "issue:43720|issue:44112", + "issue:43388|issue:43526", + "issue:45459|issue:46234", + "issue:45310|issue:45478", + "issue:43335|issue:44112", + "issue:43335|issue:45593", + "issue:45290|issue:45381", + "issue:39692|issue:44493", + "issue:28218|issue:35141", + "issue:43606|issue:44857" ], - "prepared_review_unit_hash": "2decfecdda85e977614f052febd61939aa709485f3bd0d5011ca8d406137e5f2", + "prepared_review_unit_hash": "9293f69383012f7bb181a4ecc6a2cd929c2a827e97892b8039eafbc04283f245", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 37182, - "estimated_input_tokens": 9296, - "estimated_eval_tokens": 18848 + "soft_pair_count": 34, + "serialized_chars": 38089, + "estimated_input_tokens": 9523, + "estimated_eval_tokens": 19302 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: the soft links mostly connect superficially similar transformer/model bugs, but they do not look like the same underlying issue. I would keep issue:43335 as the best canonical issue because it is the most connected concrete bug in the set; there are no PRs here.", - "confidence": 0.79, - "canonical_issue_reason": "issue:43335 has the highest soft-link connectivity in this set and is a well-scoped, concrete bug report; the rest of the cluster is mostly unrelated one-off regressions.", + "summary": "The set is mostly a collection of unrelated issue reports. Only one soft pair is a clear duplicate: the two identical Qwen3.5 Moe `from_pretrained` errors (45310/45478).", + "confidence": 0.97, + "canonical_issue_reason": "issue:45310 is the clearest canonical issue: it is a narrowly scoped, reproducible bug report and has an exact duplicate in issue:45478 with the same title and failure mode.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43335 is the best representative to anchor this cluster because it is the most central issue and still specific enough to serve as a canonical bug.", + "best_issue_reason": "issue:45310 is the best overall issue candidate because it is specific, well-scoped, and has the strongest duplicate signal in the set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44561", - "right": "issue:46234", - "accept": false, - "reason": "Both involve missing/import utility helpers, but they are different symbols and different breakages, not the same bug." - }, - { - "left": "issue:45759", - "right": "issue:46285", - "accept": false, - "reason": "Different problems: multimodal config unwrapping vs CLIP text-model structure inconsistency." - }, - { - "left": "issue:43819", - "right": "issue:44112", - "accept": false, - "reason": "One is a DAC latent/STE correctness issue; the other is a CI test failure from a stale device override." - }, - { - "left": "issue:43827", - "right": "issue:43901", - "accept": false, - "reason": "Both are docs/API regressions, but they concern different pipeline behaviors and different docs pages." - }, - { - "left": "issue:45290", - "right": "issue:46326", - "accept": false, - "reason": "Both touch chat templating, but the concrete failures differ: tool-call assistant messages vs continue_final_message for mistral_common tokenizers." - }, - { - "left": "issue:44514", - "right": "issue:44521", - "accept": false, - "reason": "Same broad API area, but one is a batched padding crash and the other is all-zero assistant masks; not the same underlying bug." - }, - { - "left": "issue:43819", - "right": "issue:45083", - "accept": false, - "reason": "Different code paths: DAC forward/latents vs qwen3_omni_moe feature-length helper behavior." - }, { "left": "issue:41720", "right": "issue:46209", "accept": false, - "reason": "Different model/hardware failures on different accelerators and inference paths." + "reason": "Both are model/runtime failures, but they concern different hardware and failure modes (cuda assert on A800 vs FP8 inference on B200)." }, { "left": "issue:44857", "right": "issue:45593", "accept": false, - "reason": "One is an AMP/CUDA loss crash; the other is missing auxiliary losses when denoising is disabled." + "reason": "Different models and bugs: AMP/CUDA crash in LwDetrImageLoss vs missing auxiliary losses in D-FINE." }, { "left": "issue:43335", "right": "issue:43550", "accept": false, - "reason": "Config construction bug vs runtime torch.compile/SDPA failure; not the same issue." + "reason": "Unrelated code paths: SwitchTransformers config layer creation vs Bamba torch.compile/SDPA failure." }, { "left": "issue:43611", "right": "issue:46402", "accept": false, - "reason": "Both are loading regressions, but they affect different model-loading mechanisms and symptoms." + "reason": "Both involve model loading, but the underlying bugs are different: `base_model_prefix` loading regression vs Sam3 text encoder weights not loaded." }, { "left": "issue:44206", "right": "issue:45584", "accept": false, - "reason": "Different regressions in different components: feature extractor arg handling vs Whisper empty-transcription generation." + "reason": "Different affected features and failures: LasrFeatureExtractor unsupported arg vs Whisper empty transcription after token alignment." }, { "left": "issue:43606", "right": "issue:45083", "accept": false, - "reason": "Device-mismatch offload bug vs unrelated qwen3_omni feature-length helper issue." + "reason": "Device mismatch with CPU offload is unrelated to the qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:37428", "right": "issue:43425", "accept": false, - "reason": "An import error for a flash-attention helper is not the same as broad Torch 2.10 incompatibility." + "reason": "Missing internal flash-attention symbol is not the same as a Torch 2.10 compatibility issue." }, { "left": "issue:42796", "right": "issue:46209", "accept": false, - "reason": "Tokenizer BPE-marker handling and FP8 inference on B200 are unrelated problems." + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are distinct problems." }, { "left": "issue:44792", "right": "issue:45563", "accept": false, - "reason": "A Janus model test failure and a stale warning in paged generate() are different defects." + "reason": "A failing image-generation test is unrelated to a stale warning in paged generate()." }, { "left": "issue:44448", "right": "issue:44743", "accept": false, - "reason": "Pegasus output mismatch and Qwen recurrent-state reset are different model behaviors/code paths." + "reason": "Different model behaviors: Pegasus output regression vs recurrent states resetting with cache in qwen3 modular code." }, { "left": "issue:43335", "right": "issue:43653", "accept": false, - "reason": "Sparse-layer config creation and BigBirdTokenizer special-token registration are unrelated." + "reason": "SwitchTransformers sparse-layer config bug and BigBirdTokenizer special-token bug are separate components and failure modes." }, { "left": "issue:45200", "right": "issue:45588", "accept": false, - "reason": "Different bugs in different subsystems: Gemma4 token-type defaults vs flash_attention aux-state handling." + "reason": "Gemma4 text-only mm token IDs issue is unrelated to the flash_attention `s_aux=None` crash." }, { "left": "issue:45127", "right": "issue:45245", "accept": false, - "reason": "LoRA merge collapse with tied embeddings is unrelated to the category-count limit error." + "reason": "LoRA/extended vocabulary collapse is unrelated to the category-count runtime error." }, { "left": "issue:43550", "right": "issue:44112", "accept": false, - "reason": "A torch.compile/SDPA model bug and a CI test issue do not look like the same underlying defect." + "reason": "Bamba torch.compile/SDPA failure and GraniteSpeech CI device-override flake are unrelated." }, { "left": "issue:44855", "right": "issue:45083", "accept": false, - "reason": "Import-time JIT parsing error vs qwen3_omni helper behavior; different code paths." + "reason": "Python 3.13 import/torch.jit parsing issue is unrelated to qwen3_omni_moe feature-length helper behavior." }, { "left": "issue:37428", "right": "issue:44351", "accept": false, - "reason": "Flash-attention helper import failure and HybridCache import failure are different missing-symbol issues." + "reason": "Missing flash-attention helper import and missing `HybridCache` import are distinct API regressions." }, { "left": "issue:43335", "right": "issue:43504", "accept": false, - "reason": "Different bugs: SwitchTransformers sparse-layer creation vs BEiT legacy-field loading." + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to Beit semantic segmentation pretrained loading with a legacy field." }, { "left": "issue:43819", "right": "issue:44493", "accept": false, - "reason": "A DAC/STE mismatch is not the same as unexpected position-id key warnings." + "reason": "DAC latent/forward mismatch is unrelated to widespread unexpected position-id keys." }, { "left": "issue:43278", "right": "issue:43825", "accept": false, - "reason": "Embedding dtype drift and an incorrect pipeline error message are unrelated regressions." + "reason": "Embedding dtype drift between train/eval is unrelated to a pipeline error-message regression." }, { "left": "issue:44206", "right": "issue:45823", "accept": false, - "reason": "Different failure modes in different models: unsupported feature-extractor arg vs PLE device mismatch." + "reason": "Different models and failure modes: LasrFeatureExtractor crash vs Gemma4 PLE device mismatch." }, { "left": "issue:29942", "right": "issue:32090", "accept": false, - "reason": "Flash Attention test failures and Trainer GPU broadcast NoneType errors are unrelated." + "reason": "Flash Attention 2 test failures and Trainer `_gpu_broadcast_one` NoneType errors are separate issues." }, { "left": "issue:39692", "right": "issue:43526", "accept": false, - "reason": "A SigLIP2 docs example bug and a BeitImageProcessorFast label-reduction bug are separate issues." + "reason": "Documentation example errors for SigLIP2 are unrelated to BeitImageProcessorFast label reduction behavior." }, { "left": "issue:45588", "right": "issue:45859", "accept": false, - "reason": "Both involve offload/attention-related code, but one crashes on s_aux=None and the other is a CPU-offload incompatibility in a rotary embedding." + "reason": "Both touch Qwen3-related code, but one is a flash-attention sink crash and the other is CPU-offload incompatibility in rotary embedding; not the same bug." }, { "left": "issue:43335", "right": "issue:45083", "accept": false, - "reason": "Sparse-layer config creation and qwen3_omni feature-length helper behavior are different bugs." + "reason": "SwitchTransformers config bug and qwen3_omni_moe helper behavior are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading and GraniteSpeech stale CI override are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Metric-gathering label truncation and Beit label reduction bug affect different subsystems and data flows." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Tokenizer error masking with missing protobuf and a missing import-utils availability helper are unrelated." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Exact duplicate reports: same title, same Qwen3.5 Moe `from_pretrained` failure, same scope." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformers config and GraniteSpeech CI test flake are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug is unrelated to D-FINE auxiliary-loss handling." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat template tool-call crash and qwen2.5-vl video position-id bug are different code paths." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 docs/example errors are unrelated to unexpected position-id keys across models." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer added-space behavior and output-embedding reinitialization are unrelated bugs." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU offload device mismatch and float16 AMP crash are different model/runtime failures." } ] }, - "evaluator_result": null, - "evaluator_used": false, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The 45310/45478 duplicate call is supported by the identical titles, and the other soft pairs are rejected without overstating similarity." + }, + "evaluator_used": true, "retried": false, - "accepted_nontrivial_soft_edge": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -8794,29 +8783,31 @@ "cluster_id": "cluster-10000-459", "nodes": [ "issue:27957", - "issue:28218", + "issue:28282", "issue:29127", "issue:29942", "issue:32090", + "issue:33357", "issue:34689", "issue:35141", + "issue:36010", + "issue:37428", "issue:38175", - "issue:39692", "issue:40444", "issue:42490", - "issue:43335", - "issue:43388", + "issue:43278", + "issue:43381", + "issue:43412", "issue:43526", "issue:43550", "issue:43575", - "issue:43606", "issue:43701", - "issue:43720", + "issue:43746", "issue:43854", "issue:43927", - "issue:44112", "issue:44360", "issue:44448", + "issue:44466", "issue:44493", "issue:44561", "issue:44568", @@ -8826,33 +8817,21 @@ "issue:44977", "issue:45072", "issue:45083", + "issue:45198", "issue:45229", - "issue:45290", - "issue:45310", - "issue:45381", - "issue:45459", - "issue:45478", + "issue:45372", "issue:45491", "issue:45563", "issue:45588", - "issue:45593", "issue:45698", + "issue:45803", + "issue:45823", "issue:45920", "issue:46192", - "issue:46234", + "issue:46209", "issue:46399" ], "soft_pairs": [ - "issue:43720|issue:44112", - "issue:43388|issue:43526", - "issue:45459|issue:46234", - "issue:45310|issue:45478", - "issue:43335|issue:44112", - "issue:43335|issue:45593", - "issue:45290|issue:45381", - "issue:39692|issue:44493", - "issue:28218|issue:35141", - "issue:43606|issue:44857", "issue:44448|issue:44779", "issue:42490|issue:43854", "issue:44360|issue:44561", @@ -8874,232 +8853,238 @@ "issue:27957|issue:34689", "issue:44857|issue:45588", "issue:44977|issue:45563", - "issue:44568|issue:45491" + "issue:44568|issue:45491", + "issue:28282|issue:33357", + "issue:43278|issue:43381", + "issue:43412|issue:45588", + "issue:45372|issue:45588", + "issue:45588|issue:45823", + "issue:36010|issue:37428", + "issue:43746|issue:46209", + "issue:44448|issue:44466", + "issue:45198|issue:45803" ], - "prepared_review_unit_hash": "c7a22d52c0c03e20047e57438a11d0e3827a55c77808aa4b352ba93be221e52f", + "prepared_review_unit_hash": "c1f00dc15778c440f8311ff1cba0a621fba6b42cb99c0b90310b6d1e8825f269", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 37485, - "estimated_input_tokens": 9372, - "estimated_eval_tokens": 19000 + "soft_pair_count": 31, + "serialized_chars": 37214, + "estimated_input_tokens": 9304, + "estimated_eval_tokens": 18864 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Only one soft duplicate looks strong: the two Qwen3.5-MoE from_pretrained error issues are effectively the same report. The rest are different bugs touching different models, APIs, or failure modes, so I would keep them separate.", - "confidence": 0.91, - "canonical_issue_reason": "issue:45310 is the best canonical issue: it is the earlier report and is an exact title match for issue:45478, so it cleanly represents the underlying bug.", + "summary": "These issues are mostly unrelated. The soft pairs share broad surface area (tokenizers, loading, flash-attn, mixed precision, etc.) but not the same concrete bug or fix path, so none of the candidate duplicates should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45920 is the broadest, most central user-facing regression here: it affects AutoTokenizer across multiple models and is likely to subsume related tokenizer reports better than the narrower issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the best overall issue in this set because it is the clearest canonical duplicate target and has the earliest report of the shared Qwen3.5-MoE from_pretrained failure.", + "best_issue_reason": "issue:45920 is the strongest representative issue because it is open, impacts several model families, and describes a high-level regression with clear user impact.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43720", - "right": "issue:44112", - "accept": false, - "reason": "Different subsystems and failures: BitNet accelerate loading vs a GraniteSpeech CI device-override test." - }, - { - "left": "issue:43388", - "right": "issue:43526", - "accept": false, - "reason": "Both mention labels, but the bugs differ: last-batch metric gathering vs BeitImageProcessorFast label reduction." - }, - { - "left": "issue:45459", - "right": "issue:46234", - "accept": false, - "reason": "Unrelated issues: tokenizer error masking with missing protobuf vs a missing import_utils availability helper." - }, - { - "left": "issue:45310", - "right": "issue:45478", - "accept": true, - "reason": "Same underlying bug: identical Qwen3.5-MoE from_pretrained failure report, just duplicated later." - }, - { - "left": "issue:43335", - "right": "issue:44112", - "accept": false, - "reason": "Different models and problems: SwitchTransformers sparse-layer config vs GraniteSpeech CI failure." - }, - { - "left": "issue:43335", - "right": "issue:45593", - "accept": false, - "reason": "Not the same bug; one is SwitchTransformers config construction, the other is D-FINE auxiliary loss behavior." - }, - { - "left": "issue:45290", - "right": "issue:45381", - "accept": false, - "reason": "Both are Qwen-related, but the bugs are different: chat template tool-call crash vs wrong vision_position_ids for video input." - }, - { - "left": "issue:39692", - "right": "issue:44493", - "accept": false, - "reason": "Different kinds of regressions: SigLIP2 doc example errors vs unexpected position-id keys across many models." - }, - { - "left": "issue:28218", - "right": "issue:35141", - "accept": false, - "reason": "Tokenizer spacing bug vs embedding reinitialization bug; same broad area but not the same issue." - }, - { - "left": "issue:43606", - "right": "issue:44857", - "accept": false, - "reason": "Different failure modes and code paths: CPU offload device mismatch vs float16 AMP crash in LwDetrImageLoss." - }, { "left": "issue:44448", "right": "issue:44779", "accept": false, - "reason": "Both are version regressions, but for different models and symptoms: Pegasus output drift vs DeepSeek tokenizer errors." + "reason": "Both are tokenizer/output regressions, but they affect different models and likely different code paths." }, { "left": "issue:42490", "right": "issue:43854", "accept": false, - "reason": "Unrelated loading issues: trailing-slash path warning vs GLM-4.7-Flash unit-test loading failure." + "reason": "A local-path import warning is not the same bug as failing to load a specific model in tests." }, { "left": "issue:44360", "right": "issue:44561", "accept": false, - "reason": "Different topics: DSA indexer architecture comment vs a v5 compatibility break from removing is_torch_fx_available." + "reason": "DSA/ReLU discussion is unrelated to the v5 trust_remote_code import breakage." }, { "left": "issue:44857", "right": "issue:45072", "accept": false, - "reason": "Both are dtype-related, but the models and crashes differ: LwDetrImageLoss AMP vs SwitchTransformers/TimmWrapper bfloat16 mismatch." + "reason": "Both involve precision issues, but they are model-specific crashes in different subsystems and not one fixable bug." }, { "left": "issue:38175", "right": "issue:44493", "accept": false, - "reason": "Different symptoms and scope: zero probabilities in one SigLIP2 model vs unexpected position-id keys across many models." + "reason": "Unexpected zero probabilities and unexpected position-id keys are different failure modes." }, { "left": "issue:45920", "right": "issue:46192", "accept": false, - "reason": "AutoTokenizer wrong IDs is a different bug from return_tensors being ignored when text_kwargs is passed." + "reason": "Tokenizer ID regressions are not the same as return_tensors being ignored when text_kwargs is passed." }, { "left": "issue:34689", "right": "issue:35141", "accept": false, - "reason": "Different model-loading regression vs embedding reinitialization after resize; not the same underlying defect." + "reason": "Model-loading failure for Llama 3.2 Vision is unrelated to embedding resize reinitialization." }, { "left": "issue:43701", "right": "issue:45083", "accept": false, - "reason": "Resume-from-checkpoint key mismatch is unrelated to qwen3_omni_moe feature-length helper behavior." + "reason": "Checkpoint key mismatch and qwen3_omni_moe feature-length behavior are different code-path problems." }, { "left": "issue:44977", "right": "issue:45229", "accept": false, - "reason": "Completely different failures: Qwen3.5 flash-attention generation vs Gemma4 multi-GPU CUDA OOM." + "reason": "Flash-attention generation failure and multi-GPU OOM are unrelated bugs." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and different regressions: generation/output mismatch vs flash-attention behavior." }, { "left": "issue:44855", "right": "issue:46399", "accept": false, - "reason": "Both mention decorator-related errors, but one is Python 3.13 parsing of DebertaV2, the other is Qwen3.5 kernel registration." + "reason": "Both mention decorators/parsing, but one is a Python 3.13 script parse issue and the other is a missing registration decorator." }, { "left": "issue:29127", "right": "issue:32090", "accept": false, - "reason": "Different bugs: clearer LayoutLMv3 error messaging vs Trainer broadcast TypeError with NoneType." + "reason": "LayoutLMv3 box error messaging is unrelated to Trainer GPU broadcast type errors." }, { "left": "issue:43550", "right": "issue:45083", "accept": false, - "reason": "Both involve model/runtime behavior, but one is Bamba torch.compile with SDPA and the other is qwen3_omni_moe feature-length logic." + "reason": "torch.compile/SDPA failure in Bamba is not the same as qwen3_omni_moe helper behavior." }, { "left": "issue:43927", "right": "issue:45698", "accept": false, - "reason": "Different save/load bugs: DiaConfig token IDs persistence vs loading the wrong custom module after save_pretrained." + "reason": "Saving custom token IDs and loading the wrong custom module are both save/load bugs, but not the same underlying issue." }, { "left": "issue:42490", "right": "issue:43575", "accept": false, - "reason": "Different loading problems: trailing-slash import warning vs Qwen2-57B-A14B-Instruct tensor-parallel OOM." + "reason": "A trailing-slash import warning is unrelated to tensor-parallel OOM on Qwen2-57B." }, { "left": "issue:40444", "right": "issue:43526", "accept": false, - "reason": "Different domains and causes: Qwen2.5-VL iterable multi-image fine-tuning vs BeitImageProcessorFast label reduction." + "reason": "Multiple-image iterable dataset failures and BEiT reduce_labels behavior are different preprocessing bugs." }, { "left": "issue:29942", "right": "issue:35141", "accept": false, - "reason": "Flash Attention 2 test failures are unrelated to embedding reinitialization after resize." + "reason": "Flash Attention 2 test failures do not match the embedding reinit bug." }, { "left": "issue:44561", "right": "issue:44779", "accept": false, - "reason": "A v5 compatibility break from removed helper is unrelated to DeepSeek tokenizer output regressions." + "reason": "A removed import helper breaking trust_remote_code is not the same as tokenizer output corruption." }, { "left": "issue:27957", "right": "issue:34689", "accept": false, - "reason": "Different loading/runtime issues for different models: XLMRoberta Flash Attention 2 vs Llama 3.2 Vision load breakage." + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Llama 3.2 Vision model loading." }, { "left": "issue:44857", "right": "issue:45588", "accept": false, - "reason": "Different flash-attention-related crashes: loss AMP overflow vs AttributeError on s_aux=None in integration code." + "reason": "LwDetr AMP crashes and sink-less flash_attention AttributeError are different failures." }, { "left": "issue:44977", "right": "issue:45563", "accept": false, - "reason": "One is generation failure with flash-attention; the other is a stale warning in paged generate() for num_return_sequences." + "reason": "Flash-attention generation correctness and a stale warning in paged generate are not the same bug." }, { "left": "issue:44568", "right": "issue:45491", "accept": false, - "reason": "Different tokenizer/model behaviors: add_special_tokens regression vs Gemma3 NaN embeddings from mixed-length batching." + "reason": "Tokenizer special-token handling and Gemma3 NaN embeddings have different causes and fixes." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "Missing PyTorch import and MacOS bus error on a CLIP model are unrelated load/runtime issues." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "BF16-to-FP32 dtype drift in eval is not the same as gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export guards/erfinv support and flash_attention s_aux handling are different code paths." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Missing ReasoningEffort import for Gemma 4 processors is unrelated to flash_attention sink-less crashes." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Flash_attention AttributeError and Gemma4 PLE device mismatch are separate runtime bugs." + }, + { + "left": "issue:36010", + "right": "issue:37428", + "accept": false, + "reason": "Both are import errors, but they involve different removed symbols and different modules." + }, + { + "left": "issue:43746", + "right": "issue:46209", + "accept": false, + "reason": "PEFT local-checkpoint loading and FP8 inference on B200 are unrelated problems." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Different output regression for Pegasus and lm_head serialization inconsistency are not the same defect." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare except swallowing errors." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The analysis is well grounded in the packet. The canonical choice is supported by the exact title match and earlier creation date for issue:45310 vs issue:45478, and the remaining soft-edge rejections are conservative and align with the differing titles/symptoms in the packet." - }, - "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 }, @@ -9116,9 +9101,10 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ + "issue:10000", + "issue:15354", "issue:25251", "issue:26413", - "issue:28282", "issue:29127", "issue:29942", "issue:30064", @@ -9126,55 +9112,45 @@ "issue:32090", "issue:33357", "issue:34567", - "issue:36010", - "issue:37428", "issue:39692", + "issue:42490", "issue:42915", "issue:42994", "issue:43065", + "issue:43066", "issue:43278", "issue:43381", - "issue:43412", "issue:43575", "issue:43611", "issue:43653", - "issue:43746", + "issue:43701", + "issue:43873", "issue:43976", "issue:43994", "issue:44186", "issue:44206", + "issue:44265", "issue:44361", - "issue:44448", - "issue:44466", + "issue:44442", + "issue:44462", "issue:44479", + "issue:44734", "issue:44743", "issue:44871", "issue:44977", "issue:45198", "issue:45245", "issue:45372", + "issue:45406", "issue:45412", "issue:45440", "issue:45459", "issue:45563", - "issue:45588", "issue:45712", - "issue:45803", - "issue:45823", "issue:46121", - "issue:46209", "issue:46369" ], "soft_pairs": [ - "issue:28282|issue:33357", - "issue:43278|issue:43381", - "issue:43412|issue:45588", - "issue:45372|issue:45588", - "issue:45588|issue:45823", - "issue:36010|issue:37428", - "issue:43746|issue:46209", - "issue:44448|issue:44466", - "issue:45198|issue:45803", "issue:25251|issue:32090", "issue:45245|issue:45563", "issue:43653|issue:44743", @@ -9194,210 +9170,205 @@ "issue:44206|issue:44479", "issue:45372|issue:45459", "issue:43994|issue:45412", - "issue:29942|issue:30064" + "issue:29942|issue:30064", + "issue:43066|issue:44462", + "issue:10000|issue:15354", + "issue:42490|issue:43873", + "issue:44265|issue:45198", + "issue:43381|issue:43701", + "issue:44734|issue:45406", + "issue:43653|issue:44442" ], - "prepared_review_unit_hash": "f15bb1b81c3de8b26f2e584673602c46fdea69f44d51b326dee1c0fc73fd60c3", + "prepared_review_unit_hash": "582bf6d2056fe33afbc36b2e904511066d2f22cbd5117b4f2a8a5a178eba0a39", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36529, - "estimated_input_tokens": 9133, - "estimated_eval_tokens": 18522 + "soft_pair_count": 27, + "serialized_chars": 35729, + "estimated_input_tokens": 8933, + "estimated_eval_tokens": 18122 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is mostly heterogeneous false positives: several items are broadly about import errors, tokenizer regressions, or quantization issues, but none of the soft pairs look like the same concrete bug/change closely enough to merge. I would not collapse any of the candidate pairs.", - "confidence": 0.95, - "canonical_issue_reason": "issue:45372 is the most concrete, actively discussed report with a clear failure mode and enough context to serve as a weak anchor if one issue must represent the set.", + "summary": "All suggested soft pairs are false matches: they describe different bugs or changes in different subsystems (tokenizers, training, serving, quantization, model loading, docs, and memory behavior). There is no coherent duplicate cluster or single canonical artifact here.", + "confidence": 0.99, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:45372 is the best representative only in the sense that it is detailed and actionable; the rest of the cluster is too unrelated for a strong duplicate anchor.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:28282", - "right": "issue:33357", - "accept": false, - "reason": "Missing PyTorch import vs MacOS bus error in CLIP; different failure modes and code paths." - }, - { - "left": "issue:43278", - "right": "issue:43381", - "accept": false, - "reason": "Training/eval dtype drift is unrelated to gradient checkpointing being disallowed in eval mode." - }, - { - "left": "issue:43412", - "right": "issue:45588", - "accept": false, - "reason": "Executorch export guard/erfinv support vs flash-attention null-handling crash; not the same bug." - }, - { - "left": "issue:45372", - "right": "issue:45588", - "accept": false, - "reason": "Gemma 4 processor import failure vs flash-attention s_aux=None crash are unrelated." - }, - { - "left": "issue:45588", - "right": "issue:45823", - "accept": false, - "reason": "Flash-attention sink-less crash and device_map auto device mismatch are different concrete problems." - }, - { - "left": "issue:36010", - "right": "issue:37428", - "accept": false, - "reason": "Both are import errors, but they involve different missing symbols in different modules." - }, - { - "left": "issue:43746", - "right": "issue:46209", - "accept": false, - "reason": "Local checkpoint loading with PEFT adapters vs FP8 inference on B200 are unrelated." - }, - { - "left": "issue:44448", - "right": "issue:44466", - "accept": false, - "reason": "Generation output regression and lm_head serialization inconsistency are different issues." - }, - { - "left": "issue:45198", - "right": "issue:45803", - "accept": false, - "reason": "Wav2Vec2 save/tokenization failure is unrelated to FuyuBatchFeature swallowing errors in tensor conversion." - }, { "left": "issue:25251", "right": "issue:32090", "accept": false, - "reason": "Pipeline top_k nesting bug vs Trainer _gpu_broadcast_one NoneType error are unrelated." + "reason": "Different failures: pipeline output-shape regression vs Trainer broadcast TypeError." }, { "left": "issue:45245", "right": "issue:45563", "accept": false, - "reason": "Category-count runtime limit vs stale generate() warning; not the same bug." + "reason": "Unrelated: category-cardinality error vs stale warning in paged generate()." }, { "left": "issue:43653", "right": "issue:44743", "accept": false, - "reason": "BigBirdTokenizer special-token registration issue is unrelated to recurrent state reset with cache." + "reason": "BigBird tokenizer special-token bug is unrelated to Qwen recurrent-state caching." }, { "left": "issue:43065", "right": "issue:45412", "accept": false, - "reason": "Dummy Conv2d placeholder and RT-DETR memory leak are distinct problems." + "reason": "Different code paths: Sam3PixelDecoder dummy conv vs RT-DETR GC/memory leak." }, { "left": "issue:44186", "right": "issue:45198", "accept": false, - "reason": "LayoutLMv2 NER padding/truncation crash is not the same as Wav2Vec2 save/tokenization failure." + "reason": "LayoutLMv2 tokenizer NER/padding crash vs Wav2Vec2 save/tokenization failure." }, { "left": "issue:44361", "right": "issue:45198", "accept": false, - "reason": "MLukeTokenizer AttributeError and Wav2Vec2 save/tokenization failure are unrelated." + "reason": "MLukeTokenizer AttributeError is unrelated to Wav2Vec2 save_pretrained/tokenization issues." }, { "left": "issue:43611", "right": "issue:46121", "accept": false, - "reason": "base_model_prefix loading regression and rope-params JSON list TypeError are different code paths." + "reason": "Model loading with base_model_prefix is unrelated to rope-validation type handling." }, { "left": "issue:33357", "right": "issue:39692", "accept": false, - "reason": "MacOS CLIP bus error and SigLIP2 docs/quantization errors are not the same issue." + "reason": "MacOS bus error on CLIP model differs from SigLIP2 doc/example errors." }, { "left": "issue:29127", "right": "issue:34567", "accept": false, - "reason": "LayoutLMv3 error-message clarity is unrelated to TrainerState token counting." + "reason": "LayoutLMv3 box-clarity error is unrelated to TrainerState token-count tracking." }, { "left": "issue:42915", "right": "issue:42994", "accept": false, - "reason": "FP8 inference failure on Qwen3Moe and quantized model saving failure are different bugs." + "reason": "FineGrainedFP8Config model failure is not the same as quantized-model save failure." }, { "left": "issue:45459", "right": "issue:45712", "accept": false, - "reason": "Tokenizer protobuf error handling and leftover dummy classes are unrelated repository issues." + "reason": "Tokenizer exception masking vs leftover dummy classes leaking into repo checks." }, { "left": "issue:44871", "right": "issue:44977", "accept": false, - "reason": "Gemma eos_token_id mismatch and Qwen3.5 flash-attention generation regression do not share a concrete code path." + "reason": "Gemma eos_token_id mismatch is unrelated to Qwen3.5 flash-attention generation." }, { "left": "issue:43976", "right": "issue:45440", "accept": false, - "reason": "Python version compatibility break and DeepseekV3MoE divergence are different kinds of failures." + "reason": "Python-version compatibility issue is not the same as DeepseekV3MoE divergence." }, { "left": "issue:45412", "right": "issue:46369", "accept": false, - "reason": "RT-DETR GC memory leak and Sam3-litetext memory overflow are separate memory problems." + "reason": "RT-DETR object lifetime leak differs from sam3-litetext memory overflow." }, { "left": "issue:43278", "right": "issue:43575", "accept": false, - "reason": "Embedding dtype drift in eval and tensor-parallel OOM on Qwen2-57B are unrelated." + "reason": "BF16/FP32 eval dtype change is unrelated to tp-related OOM on Qwen2-57B." }, { "left": "issue:26413", "right": "issue:31501", "accept": false, - "reason": "Resume-from-checkpoint epoch_iterator failure is not the same as FineWeb training not starting." + "reason": "Checkpoint resume iterator error is not the same as FineWeb training startup failure." }, { "left": "issue:44206", "right": "issue:44479", "accept": false, - "reason": "Unsupported center arg in an extractor and Qwen video-input regression are different failures." + "reason": "Unsupported feature-extractor arg differs from video-input regression in Qwen variants." }, { "left": "issue:45372", "right": "issue:45459", "accept": false, - "reason": "ReasoningEffort import break and protobuf decode-error masking are unrelated import/error-handling bugs." + "reason": "Gemma 4 processor import breakage is unrelated to protobuf error masking in tokenizers." }, { "left": "issue:43994", "right": "issue:45412", "accept": false, - "reason": "SigLIP2 nonsensical outputs and RT-DETR memory leak are unrelated." + "reason": "SigLIP2 incorrect outputs are not a memory-release problem." }, { "left": "issue:29942", "right": "issue:30064", "accept": false, - "reason": "Flash Attention 2 test failures and void segmentation map processing are different model/test issues." + "reason": "Flash Attention 2 test failures are unrelated to void segmentation-map processing." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type and ignoring tokenizer.json are different tokenizer-load bugs." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "DistilBert model issue in German is unrelated to torch.jit GeneratorExp export failure." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Trailing-slash import warning is not the same as offloading/quantization behavior." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export + torch_compilable_check failure is unrelated to Wav2Vec2 save/tokenization failure." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Gradient-checkpointing eval limitation is different from resume_from_checkpoint key mismatch." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation indexing crash is unrelated to Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer mask-token registration bug is unrelated to AutoTokenizer loading FastSpeech2ConformerTokenizer." } ] }, @@ -9421,17 +9392,17 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:10000", - "issue:15354", "issue:27957", + "issue:29127", "issue:29942", "issue:32090", "issue:33357", + "issue:35141", "issue:41720", "issue:42490", "issue:42915", "issue:43065", - "issue:43066", + "issue:43262", "issue:43278", "issue:43329", "issue:43334", @@ -9440,7 +9411,6 @@ "issue:43526", "issue:43531", "issue:43575", - "issue:43653", "issue:43701", "issue:43756", "issue:43761", @@ -9449,35 +9419,29 @@ "issue:43931", "issue:43994", "issue:44186", - "issue:44265", + "issue:44206", "issue:44279", "issue:44387", "issue:44442", - "issue:44462", - "issue:44734", + "issue:44464", "issue:44857", "issue:44977", "issue:45072", - "issue:45198", "issue:45229", "issue:45245", "issue:45372", "issue:45406", "issue:45440", "issue:45491", + "issue:45584", + "issue:45588", "issue:45593", + "issue:45600", "issue:45834", "issue:45859", "issue:46369" ], "soft_pairs": [ - "issue:43066|issue:44462", - "issue:10000|issue:15354", - "issue:42490|issue:43873", - "issue:44265|issue:45198", - "issue:43381|issue:43701", - "issue:44734|issue:45406", - "issue:43653|issue:44442", "issue:43421|issue:45491", "issue:43873|issue:45229", "issue:43065|issue:43994", @@ -9501,222 +9465,215 @@ "issue:44977|issue:45245", "issue:43994|issue:46369", "issue:43824|issue:44279", - "issue:42915|issue:43931" + "issue:42915|issue:43931", + "issue:27957|issue:29127", + "issue:29127|issue:35141", + "issue:43262|issue:44206", + "issue:44464|issue:45584", + "issue:45588|issue:45600" ], - "prepared_review_unit_hash": "08e68e729ea736be640cb98d2ae0a56f3ec3fcdce8641d0f79c4a4e17863363a", + "prepared_review_unit_hash": "a007fa62dd03294648bc13a037dbcba6e3d686c49b94353e202f9d5890543103", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 31, - "serialized_chars": 36877, - "estimated_input_tokens": 9220, - "estimated_eval_tokens": 18696 + "soft_pair_count": 29, + "serialized_chars": 36353, + "estimated_input_tokens": 9089, + "estimated_eval_tokens": 18434 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This is not a true duplicate cluster: the soft links are all false positives across unrelated bugs, models, and code paths. The titles overlap lexically (tokenizer, flash-attn, OOM, serve, etc.) but not at the concrete failure level.", - "confidence": 0.98, - "canonical_issue_reason": "No strong canonical issue: the set is heterogeneous, with only superficial topic overlap.", + "summary": "The cluster is not a true duplicate set: the paired issues share only broad vocabulary (e.g. Flash Attention, loading, OOM, multimodal) but describe different models, code paths, and failure modes. No PRs are present; the broadest representative issue is the open quantization/offloading report.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43873 is the broadest, still-open problem statement in the set and best represents the general runtime/memory-regression theme, though the cluster overall is heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the broadest/highest-activity report in the set, but it still does not anchor the cluster well.", + "best_issue_reason": "issue:43873 has the widest scope and highest likelihood of anchoring related follow-up work among these mostly unrelated issues.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43066", - "right": "issue:44462", - "accept": false, - "reason": "Different tokenizer bugs: decoder type mismatch vs ignoring tokenizer.json." - }, - { - "left": "issue:10000", - "right": "issue:15354", - "accept": false, - "reason": "Unrelated models and failures: DistilBERT issue vs torch.jit generator-exp export error." - }, - { - "left": "issue:42490", - "right": "issue:43873", - "accept": false, - "reason": "Import warning for trailing slash path vs quantization/offloading memory behavior." - }, - { - "left": "issue:44265", - "right": "issue:45198", - "accept": false, - "reason": "Different code paths: torch.export/torch_compilable_check vs Wav2Vec2 save/tokenization." - }, - { - "left": "issue:43381", - "right": "issue:43701", - "accept": false, - "reason": "Checkpointing-in-eval bug vs resume_from_checkpoint key mismatch; not the same defect." - }, - { - "left": "issue:44734", - "right": "issue:45406", - "accept": false, - "reason": "Both are serve-related, but one is KV-cache indexing and the other is Gemma4Processor init." - }, - { - "left": "issue:43653", - "right": "issue:44442", - "accept": false, - "reason": "BigBird special-token decode bug vs AutoTokenizer loading the wrong tokenizer source." - }, { "left": "issue:43421", "right": "issue:45491", "accept": false, - "reason": "Runtime post-processor sync issue vs Gemma3 sliding-window NaNs; different failures." + "reason": "Different bugs: tokenizer post-processor update behavior vs Gemma3 sliding-window NaNs." }, { "left": "issue:43873", "right": "issue:45229", "accept": false, - "reason": "Both mention memory pressure, but one is offloading+quantization and the other is multi-GPU OOM." + "reason": "Both mention memory pressure, but one is quantization offloading behavior and the other is multi-GPU OOM for Gemma4." }, { "left": "issue:43065", "right": "issue:43994", "accept": false, - "reason": "Different models and symptoms: Sam3PixelDecoder dummy layer vs SigLIP2 nonsense outputs." + "reason": "Unrelated components: Sam3PixelDecoder dummy conv vs SigLIP2 wrong outputs with AutoModel/pipeline." }, { "left": "issue:27957", "right": "issue:29942", "accept": false, - "reason": "Flash-Attn 2 support request vs failing Flash-Attn 2 tests; not one concrete bug." + "reason": "Same Flash Attention 2 wording, but one is XLM-R support and the other is failing FA2 tests; different underlying issues." }, { "left": "issue:32090", "right": "issue:33357", "accept": false, - "reason": "Trainer broadcast NoneType error vs MacOS bus error on a CLIP model." + "reason": "Different failures: Trainer NoneType broadcast error vs MacOS bus error with a community CLIP model." }, { "left": "issue:43756", "right": "issue:45440", "accept": false, - "reason": "Different model implementations: Smollm3 RoPE-layer issue vs DeepseekV3MoE divergence." + "reason": "Different model families and bugs: Smollm3 RoPE-layer mismatch vs DeepseekV3 native implementation divergence." }, { "left": "issue:45372", "right": "issue:45406", "accept": false, - "reason": "Gemma4 import dependency error vs Gemma4Processor missing _tokenizer; different root causes." + "reason": "Gemma4 processor import dependency break vs Gemma4 processor missing _tokenizer; separate failures." }, { "left": "issue:43526", "right": "issue:45834", "accept": false, - "reason": "Beit reduce_labels bug vs Kosmos2.5 long-OCR index error; unrelated." + "reason": "Both are vision/OCR-ish, but one is BeitImageProcessorFast label reduction and the other is Kosmos2.5 long OCR index error." }, { "left": "issue:41720", "right": "issue:45229", "accept": false, - "reason": "Both are large-model memory failures, but on different models and with different triggers." + "reason": "Different root causes: Qwen3 auto device mapping cudaErrorAssert vs Gemma4 multi-GPU OOM." }, { "left": "issue:43278", "right": "issue:43329", "accept": false, - "reason": "BF16-to-FP32 eval dtype drift vs undefined video helper/merge_size in multimodal token counting." + "reason": "Unrelated bugs: dtype drift in embedding layers vs undefined variables in multimodal token counting video path." }, { "left": "issue:45440", "right": "issue:45593", "accept": false, - "reason": "DeepseekV3 implementation mismatch vs D-FINE auxiliary-loss behavior; no shared code defect." + "reason": "DeepSeekV3 implementation mismatch is unrelated to D-FINE auxiliary-loss handling." }, { "left": "issue:43575", "right": "issue:43701", "accept": false, - "reason": "Qwen2 tensor-parallel OOM vs checkpoint resume key mismatch; unrelated." + "reason": "Different problems: tensor-parallel OOM on Qwen2-57B vs resume_from_checkpoint key mismatch." }, { "left": "issue:43526", "right": "issue:44186", "accept": false, - "reason": "Different tokenizer/image-processor bugs with no shared failure path." + "reason": "Both involve processors/tokenizers, but BeitImageProcessorFast label reduction and LayoutLMv2 NER padding crashes are different bugs." }, { "left": "issue:42490", "right": "issue:43381", "accept": false, - "reason": "Trailing-slash import warning vs gradient-checkpointing eval-mode restriction." + "reason": "Trailing-slash local model import warning is unrelated to gradient checkpointing in eval mode." }, { "left": "issue:43761", "right": "issue:45859", "accept": false, - "reason": "CLIPVisionModel hidden_states regression vs Qwen3 MoE CPU-offload incompatibility." + "reason": "Different code paths: CLIPVisionModel hidden_states regression vs Qwen3.5Moe CPU offload incompatibility." }, { "left": "issue:43575", "right": "issue:44387", "accept": false, - "reason": "Both are OOM-related, but one is model-parallel loading and the other is int4 reserved-memory regression." + "reason": "Both are memory-related, but one is model-loading OOM and the other is increased CUDA reserved memory under int4 quantization." }, { "left": "issue:43526", "right": "issue:44857", "accept": false, - "reason": "Beit label reduction bug vs LwDetrImageLoss AMP crash; different subsystems." + "reason": "Different image-processing bugs: label reduction vs image loss crashing under float16 AMP on CUDA." }, { "left": "issue:27957", "right": "issue:33357", "accept": false, - "reason": "XLMRoberta Flash-Attn 2 issue vs CLIP MacOS bus error; unrelated." + "reason": "No shared underlying bug: Flash Attention 2 support for XLM-R vs MacOS bus error loading a CLIP model." }, { "left": "issue:43334", "right": "issue:43531", "accept": false, - "reason": "Qwen3-VL load error due to pad_token_id vs Qwen3-MoE sliding-window behavior." + "reason": "Different Qwen3-VL vs Qwen3-MoE loading/runtime issues; one is missing pad_token_id, the other sliding_window behavior." }, { "left": "issue:44442", "right": "issue:45072", "accept": false, - "reason": "Tokenizer loading problem vs bfloat16 dtype mismatch; different bugs." + "reason": "Different tokenizer/model problems: FastSpeech2ConformerTokenizer loading vs dtype mismatch in SwitchTransformers/TimmWrapperModel." }, { "left": "issue:44977", "right": "issue:45245", "accept": false, - "reason": "Flash-attention generation issue vs category-count limit RuntimeError; unrelated." + "reason": "Qwen3.5 flash-attention generation bug is unrelated to category-count runtime limit." }, { "left": "issue:43994", "right": "issue:46369", "accept": false, - "reason": "SigLIP2 output corruption vs Sam3-litetext memory overflow." + "reason": "Different bugs: SigLIP2 output quality issue vs Sam3-lite text memory overflow." }, { "left": "issue:43824", "right": "issue:44279", "accept": false, - "reason": "Specific Qwen2_5_VL import failure vs vague dependency issue; not the same problem." + "reason": "Importing Qwen2.5-VL is a specific API exposure issue; dependency issue is broader and not the same bug." }, { "left": "issue:42915", "right": "issue:43931", "accept": false, - "reason": "Qwen3Moe FP8 config failure vs Qwen3-VL weight-shape mismatch; different models and errors." + "reason": "Different Qwen3-Moe FP8 failure vs Qwen3-VL shape-mismatch loading error." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "XLM-R Flash Attention 2 support is unrelated to LayoutLMv3 box-validation error messaging." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Different subsystems: LayoutLMv3 error clarity vs post_init reinitializing tied output embeddings." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting is unrelated to LasrFeatureExtractor center-arg crash." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation with compiled forward is a different failure mode from Whisper empty-transcription handling after token alignment." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "Flash-attention sink-less-model crash is unrelated to auto_mappings referencing removed Sam3LiteText configs." } ] }, @@ -9742,17 +9699,14 @@ "nodes": [ "issue:27957", "issue:28218", - "issue:29127", "issue:32090", "issue:33290", "issue:34567", "issue:34689", - "issue:35141", "issue:36010", "issue:42175", "issue:42796", "issue:42907", - "issue:43262", "issue:43316", "issue:43381", "issue:43493", @@ -9761,28 +9715,30 @@ "issue:43611", "issue:43646", "issue:43720", + "issue:43761", "issue:43827", "issue:43873", "issue:43931", "issue:44062", "issue:44077", - "issue:44206", + "issue:44188", "issue:44442", "issue:44464", "issue:44743", "issue:44811", + "issue:44855", "issue:44871", "issue:44898", "issue:44964", "issue:44977", + "issue:45137", "issue:45216", "issue:45290", "issue:45440", "issue:45491", "issue:45584", - "issue:45588", - "issue:45600", "issue:45698", + "issue:45759", "issue:45803", "issue:45907", "issue:46121", @@ -9790,11 +9746,6 @@ "issue:46399" ], "soft_pairs": [ - "issue:27957|issue:29127", - "issue:29127|issue:35141", - "issue:43262|issue:44206", - "issue:44464|issue:45584", - "issue:45588|issue:45600", "issue:44743|issue:44811", "issue:27957|issue:33290", "issue:43646|issue:44077", @@ -9819,216 +9770,207 @@ "issue:44442|issue:45803", "issue:42175|issue:46234", "issue:42796|issue:45290", - "issue:43873|issue:44871" + "issue:43873|issue:44871", + "issue:43316|issue:45759", + "issue:43761|issue:44188", + "issue:44855|issue:45137" ], - "prepared_review_unit_hash": "8b5960db47ff7b1c11af36e69be00639f16b4009cd9fe82ceaad8774c24470c4", + "prepared_review_unit_hash": "d7c9239a8bbbe2571c953df2aef4e152c7c4d6f9be76f53ae3bd5645c0ad8a97", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 30, - "serialized_chars": 36804, - "estimated_input_tokens": 9201, - "estimated_eval_tokens": 18658 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35624, + "estimated_input_tokens": 8906, + "estimated_eval_tokens": 18068 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is a loose collection of Transformers issues with superficial title overlap (tokenizers, loading, generation, config, training). The soft links are mostly same-subsystem similarity, not the same underlying bug or mergeable change, so I would not collapse any of them as duplicates.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 45698 is the broadest and most active open regression in the set, centered on save/load behavior, so it is the best single representative if one must be chosen.", - "canonical_pr_reason": null, - "best_issue_reason": "Issue 45698 has the strongest triage value: open, well-discussed, and affecting a core serialization path that can surface in multiple downstream workflows.", - "best_pr_reason": null, + "summary": "The items are a heterogeneous mix of unrelated issues; none of the proposed soft pairs look like the same underlying bug or change, so all soft edges should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out: this set spans unrelated tokenizer, training, loading, quantization, generation, and documentation bugs, so there is no cohesive duplicate cluster to collapse onto one issue.", + "canonical_pr_reason": "No PRs are present in the input.", + "best_issue_reason": "No best issue among these for deduplication purposes; the candidates are too diverse and do not share a common code path or fix.", + "best_pr_reason": "No PRs are present in the input.", "soft_edge_verdicts": [ - { - "left": "issue:27957", - "right": "issue:29127", - "accept": false, - "reason": "Flash Attention 2 support for XLM-Roberta and a LayoutLMv3 error-message issue are unrelated bugs in different models and code paths." - }, - { - "left": "issue:29127", - "right": "issue:35141", - "accept": false, - "reason": "One is about clearer validation errors for boxes; the other is about output embeddings being reinitialized after resize. Different failures." - }, - { - "left": "issue:43262", - "right": "issue:44206", - "accept": false, - "reason": "Both touch processor/feature-extractor behavior, but one is a wrong default sampling rate and the other is an unsupported argument crash. Not the same bug." - }, - { - "left": "issue:44464", - "right": "issue:45584", - "accept": false, - "reason": "Chunked generation with compiled forward and Whisper empty-transcription generation are different generation failures with different triggers." - }, - { - "left": "issue:45588", - "right": "issue:45600", - "accept": false, - "reason": "Flash-attention sink handling and stale auto_mappings config references are separate problems: runtime kernel crash vs CI/import mapping breakage." - }, { "left": "issue:44743", "right": "issue:44811", "accept": false, - "reason": "Recurrent state reset in Qwen3.5 and Whisper batch_decode ignoring skip_special_tokens are unrelated logic bugs." + "reason": "Different components and failures: Qwen recurrent-state reset in cache handling vs Whisper batch decode token skipping." }, { "left": "issue:27957", "right": "issue:33290", "accept": false, - "reason": "Model support for Flash Attention 2 and DeepSpeed/Adafactor OOM are entirely different issues." + "reason": "Flash Attention 2 model support and a Deepspeed/Adafactor OOM are unrelated bugs." }, { "left": "issue:43646", "right": "issue:44077", "accept": false, - "reason": "Custom model initialization regression and patchtsmixer post_init allowance are different API/validation concerns, not one underlying bug." + "reason": "Both concern initialization, but one is a broad Transformers 5 custom-model init regression and the other is a specific patchtsmixer post_init policy issue." }, { "left": "issue:44977", "right": "issue:45584", "accept": false, - "reason": "Both affect generation, but Qwen3.5 flash-attention behavior and Whisper empty-transcription handling are different model-specific failures." + "reason": "Both are generation failures, but they affect different models and code paths: Qwen3.5 flash-attention vs Whisper empty-transcription handling." }, { "left": "issue:28218", "right": "issue:32090", "accept": false, - "reason": "Tokenizer added-space behavior and Trainer _gpu_broadcast_one NoneType errors are unrelated tokenizer vs distributed-training bugs." + "reason": "Tokenizer added-token spacing is unrelated to the Trainer/_gpu_broadcast_one NoneType error." }, { "left": "issue:33290", "right": "issue:34567", "accept": false, - "reason": "Adafactor OOM in DeepSpeed and num_input_tokens_seen not updating are unrelated training issues." + "reason": "An OOM with Adafactor in Deepspeed is not the same as TrainerState token counters not updating." }, { "left": "issue:44062", "right": "issue:45907", "accept": false, - "reason": "AddedToken constructor argument conflict and list_repo_files revision forwarding are different APIs and failure modes." + "reason": "AddedToken constructor argument collision and list_repo_files revision forwarding are unrelated API bugs." }, { "left": "issue:44464", "right": "issue:45491", "accept": false, - "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaN embeddings from sliding-window attention are distinct problems in different parts of the stack." + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaN embeddings from sliding-window attention are different failure modes." }, { "left": "issue:44442", "right": "issue:44898", "accept": false, - "reason": "Tokenizer loading failure and Perceiver non-default resolution failure are both model-specific, but not the same code path or fix." + "reason": "Tokenizer loading failure and Perceiver image classification resolution handling are unrelated subsystems." }, { "left": "issue:43381", "right": "issue:43873", "accept": false, - "reason": "Gradient checkpointing in eval mode and quantization offloading are separate training/runtime behaviors." + "reason": "Gradient checkpointing in eval mode and quantization offloading are different behavioral issues." }, { "left": "issue:46121", "right": "issue:46399", "accept": false, - "reason": "Rope-validation TypeError from a JSON-loaded list and a missing decorator for Qwen3.5 rotary registration are unrelated bugs." + "reason": "A JSON-list type error in rope validation is unrelated to Qwen3.5 rotary-embedding registration under use_kernels=True." }, { "left": "issue:43316", "right": "issue:43525", "accept": false, - "reason": "Gemma3TextConfig API inconsistency and missing pad_token_id on Llama4Config are different model config defects." + "reason": "Gemma3TextConfig API mismatch and Llama4Config missing pad_token_id are separate config bugs." }, { "left": "issue:44964", "right": "issue:45440", "accept": false, - "reason": "Phi-4 multimodal loading regression and DeepSeekV3MoE divergence from remote implementation are different model-specific issues." + "reason": "Phi-4 multimodal model loading regression and DeepseekV3MoE divergence from remote implementation do not share the same fix." }, { "left": "issue:34689", "right": "issue:43611", "accept": false, - "reason": "Llama 3.2 vision loading breakage and base_model_prefix loading regression both involve loading, but the root causes are different." + "reason": "Both mention model loading breakage, but one is a version-specific Llama 3.2 vision regression and the other is a base_model_prefix-related Transformers 5 regression." }, { "left": "issue:32090", "right": "issue:33290", "accept": false, - "reason": "Trainer broadcast TypeError and DeepSpeed Adafactor OOM are unrelated runtime failures." + "reason": "Trainer broadcast type handling and Deepspeed Adafactor OOM are unrelated training/runtime bugs." }, { "left": "issue:43720", "right": "issue:44442", "accept": false, - "reason": "BitNet packed-weight unpacking during accelerate loading and tokenizer loading failure are different subsystems and bugs." + "reason": "BitNet packed-weight loading under accelerate and FastSpeech2ConformerTokenizer loading are entirely different code paths." }, { "left": "issue:43931", "right": "issue:45216", "accept": false, - "reason": "Qwen3-VL weight-shape mismatch on load and Qwen3.5 save_pretrained regression are both model-family issues, but not the same concrete bug." + "reason": "Weight-shape mismatch when loading Qwen3-VL is not the same as save_pretrained producing an incorrect checkpoint." }, { "left": "issue:34689", "right": "issue:36010", "accept": false, - "reason": "A model-loading regression and a GenerationMixin import-path issue are unrelated." + "reason": "These are separate import/load regressions with different failure points and APIs." }, { "left": "issue:43493", "right": "issue:43827", "accept": false, - "reason": "SigLIP2 implementation divergence and docs still referencing pipeline() are not the same issue; one is model behavior, the other documentation cleanup." + "reason": "SigLIP2 implementation mismatch and docs still referencing pipeline() are unrelated." }, { "left": "issue:43550", "right": "issue:44743", "accept": false, - "reason": "Bamba torch.compile/SDPA failure and Qwen3.5 recurrent state reset are different model/runtime bugs." + "reason": "A torch.compile/SDPA failure in Bamba is unrelated to Qwen recurrent-state reset logic." }, { "left": "issue:42907", "right": "issue:45698", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals and loading the wrong custom module after save_pretrained are both save/load themed, but they are not the same defect." + "reason": "Dequantized save failure and wrong custom module loading after save_pretrained are distinct serialization problems." }, { "left": "issue:44442", "right": "issue:45803", "accept": false, - "reason": "Tokenizer load failure and Fuyu tensor conversion swallowing exceptions are unrelated bugs." + "reason": "Tokenizer loading and a bare-except swallowing errors in FuyuBatchFeature.convert_to_tensors() are not the same bug." }, { "left": "issue:42175", "right": "issue:46234", "accept": false, - "reason": "Missing TensorFlow backend in a torch-only install and a missing import_utils symbol are different packaging/import issues." + "reason": "Missing TensorFlow in a torch extra install and a missing import_utils helper are unrelated packaging/API issues." }, { "left": "issue:42796", "right": "issue:45290", "accept": false, - "reason": "Ministral BPE marker handling and chat_template crashes on tool-call assistant messages are both tokenizer-related, but they affect different paths and inputs." + "reason": "Ministral BPE marker handling and chat-template crashes on tool-call messages are different tokenizer behaviors." }, { "left": "issue:43873", "right": "issue:44871", "accept": false, - "reason": "Quantization/offloading behavior and Gemma3 eos_token_id mismatch are unrelated." + "reason": "Quantization offloading and Gemma-3 eos_token_id inconsistency are separate configuration/runtime bugs." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Qwen multimodal text_config unwrapping are different config/model-construction issues." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and attention-kernel divergence under torch.compile are unrelated." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 torch.jit parsing/import error and DeepSpeed ZeRO3 deque underflow are different failures." } ] }, @@ -10073,11 +10015,10 @@ "issue:43701", "issue:43716", "issue:43742", - "issue:43761", "issue:43825", "issue:43931", "issue:44077", - "issue:44188", + "issue:44162", "issue:44190", "issue:44464", "issue:44479", @@ -10085,30 +10026,29 @@ "issue:44514", "issue:44568", "issue:44610", - "issue:44855", + "issue:44792", "issue:44871", "issue:44913", "issue:44987", - "issue:45137", "issue:45479", + "issue:45584", "issue:45632", - "issue:45759", "issue:45803", "issue:46178", "issue:46202", + "issue:46234", + "issue:46291", "issue:46365", "issue:46369", "issue:46421", "issue:46424" ], "soft_pairs": [ - "issue:43316|issue:45759", - "issue:43761|issue:44188", - "issue:44855|issue:45137", "issue:44190|issue:44496", "issue:44464|issue:46421", "issue:44987|issue:46178", "issue:43643|issue:44913", + "issue:46234|issue:46291", "issue:44077|issue:44479", "issue:44568|issue:45479", "issue:43295|issue:43825", @@ -10128,198 +10068,200 @@ "issue:42175|issue:43316", "issue:43295|issue:44871", "issue:45803|issue:46365", - "issue:43540|issue:44514" + "issue:43540|issue:44514", + "issue:44871|issue:45584", + "issue:44162|issue:44792" ], - "prepared_review_unit_hash": "a58eaa9f438d5ef8159fcac45fc74d73c6cd32703be8b85bfdcb8bee600081f6", + "prepared_review_unit_hash": "4fd33aae28d7f1db2114a705e4aab628e5c856da9cf5ee3c2f8a97541efb77d1", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 27, - "serialized_chars": 35875, - "estimated_input_tokens": 8969, - "estimated_eval_tokens": 18194 + "serialized_chars": 35741, + "estimated_input_tokens": 8936, + "estimated_eval_tokens": 18128 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: the soft-paired issues span unrelated regressions in loading, generation, tokenization, compilation, deepspeed, and processor behavior. I found no credible duplicate pairs and no meaningful single canonical issue for the set.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue stands out because the items are not a duplicate cluster; they cover many unrelated bug classes and model-specific failures.", + "summary": "This cluster is mostly a grab bag of unrelated Transformers issues: model loading regressions, processor/tokenizer bugs, backend/import problems, and a few compile/cache/training edge cases. None of the soft pairs look like the same underlying bug or change, so I would reject all of them as duplicates.", + "confidence": 0.72, + "canonical_issue_reason": "If forced to pick one anchor, issue 43611 is the broadest user-facing regression in the set: a Transformers 5.0 model-loading break that could affect many models, unlike the more model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "No single best issue for deduplication: the set is too diverse to anchor on one representative bug without misleading triage.", + "best_issue_reason": "Issue 43611 is the best representative issue for this cluster because it is a clear, high-impact regression and more general than the surrounding model-specific or narrow API problems.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43316", - "right": "issue:45759", - "accept": false, - "reason": "Different APIs and failure modes: Gemma3TextConfig configuration discrepancy vs AutoModelForCausalLM.from_config not unwrapping text_config for composite Qwen multimodal configs." - }, - { - "left": "issue:43761", - "right": "issue:44188", - "accept": false, - "reason": "CLIPVisionModel.forward missing hidden_states is a model output bug; allow_is_bidirectional_skip on torch.compile is an attention-kernel divergence issue." - }, - { - "left": "issue:44855", - "right": "issue:45137", - "accept": false, - "reason": "Python 3.13 torch.jit parsing/indentation error vs DeepSpeed ZeRO3 deque underflow; unrelated subsystems and symptoms." - }, { "left": "issue:44190", "right": "issue:44496", "accept": false, - "reason": "Local dataset loading script failure is unrelated to an unrecognized model/config loading error for a specific checkpoint." + "reason": "Local dataset loading failure and unrecognized-model/config loading failure are different bugs in different code paths." }, { "left": "issue:44464", "right": "issue:46421", "accept": false, - "reason": "Both mention chunked generation/prefill, but one is inconsistent outputs under compiled forward while the other is spurious recompilation from cache lazy init; not the same bug." + "reason": "Both mention chunking/compiled execution, but one is about inconsistent outputs while the other is about cache lazy init causing recompilation." }, { "left": "issue:44987", "right": "issue:46178", "accept": false, - "reason": "Transformers>=5.1 loading of one model vs Trainer import failing because of datasets/backend resolution; different code paths and root causes." + "reason": "A model-loading regression for one repo and a backend-mapping error for tensorflow_text are unrelated failures." }, { "left": "issue:43643", "right": "issue:44913", "accept": false, - "reason": "trust_remote_code missing returned config fields is distinct from GPTNeoXConfig rotary_pct not persisting on reload." + "reason": "Both are config-related, but one is missing fields with trust_remote_code and the other is rotary_pct not persisting on reload." + }, + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Missing import_utils helper and kernels version/revision import failure are different dependency/API issues." }, { "left": "issue:44077", "right": "issue:44479", "accept": false, - "reason": "patchtsmixer post_init allowance and Qwen video-input regression are different model/regression classes." + "reason": "PatchTSMixer post_init gating and Qwen video-input regression are unrelated model-specific bugs." }, { "left": "issue:44568", "right": "issue:45479", "accept": false, - "reason": "Tokenizer add_special_tokens regression is not the same as single-label classification with num_labels=1 producing zero loss." + "reason": "Tokenizer special-token behavior and sequence-classification zero-loss behavior are different concrete bugs." }, { "left": "issue:43295", "right": "issue:43825", "accept": false, - "reason": "Custom processor.tokenizer/image handling regression is unrelated to an incorrect pipeline translation error message." + "reason": "One is a processor/tokenizer API regression for custom model code; the other is an error-message issue in pipeline() about translation support." }, { "left": "issue:43611", "right": "issue:43931", "accept": false, - "reason": "base_model_prefix loading break and Qwen3-VL weight-shape mismatch are different model-loading failures." + "reason": "Both are loading failures, but one is a base_model_prefix regression and the other is a shape mismatch for a specific Qwen3-VL model." }, { "left": "issue:45632", "right": "issue:46424", "accept": false, - "reason": "trust_remote_code cache-path collision for local models is unrelated to max_cache_length only applying to Quantized Cache." + "reason": "Cache-path collisions for trust_remote_code and max_cache_length being ignored except for quantized cache are different cache behaviors." }, { "left": "issue:43550", "right": "issue:43701", "accept": false, - "reason": "torch.compile + SDPA failure in Bamba is unrelated to resume_from_checkpoint key mismatch." + "reason": "torch.compile/SDPA failure and resume_from_checkpoint key mismatch are unrelated." }, { "left": "issue:25251", "right": "issue:28282", "accept": false, - "reason": "Pipeline top_k changing output structure is unrelated to AutoModel failing because PyTorch is missing." + "reason": "Pipeline output nesting with top_k and missing PyTorch ImportError are not the same bug." }, { "left": "issue:43606", "right": "issue:43701", "accept": false, - "reason": "CPU offload device mismatch and checkpoint key mismatch are different bugs with no shared code path." + "reason": "CPU offload device mismatch and checkpoint key mismatch are unrelated training/runtime problems." }, { "left": "issue:33290", "right": "issue:34689", "accept": false, - "reason": "Adafactor/DeepSpeed OOM is not the same as Llama 3.2 Vision loading break." + "reason": "AdaFactor OOM in DeepSpeed and Llama 3.2 Vision loading break are different regressions." }, { "left": "issue:42617", "right": "issue:43575", "accept": false, - "reason": "3d_parallel.py runtime failure and tp-load OOM for Qwen2-57B are separate issues." + "reason": "A script execution problem and a tp-load OOM on a specific model are not the same issue." }, { "left": "issue:33290", "right": "issue:35141", "accept": false, - "reason": "Optimizer OOM in DeepSpeed is unrelated to output embedding reinitialization during post_init." + "reason": "Optimizer memory blowup and embedding reinitialization on resize are unrelated bugs." }, { "left": "issue:43575", "right": "issue:43742", "accept": false, - "reason": "Tensor-parallel OOM on a large Qwen model and a key error loading MobileLLM-125M are not the same bug." + "reason": "Both are model-loading complaints, but one is OOM under tensor parallelism and the other is a key error for a different model." }, { "left": "issue:27957", "right": "issue:32090", "accept": false, - "reason": "Flash Attention 2 support for XLMRoberta and NoneType broadcast errors in Trainer are unrelated." + "reason": "Flash Attention 2 support for XLM-Roberta and a Trainer broadcast TypeError are unrelated." }, { "left": "issue:43316", "right": "issue:43716", "accept": false, - "reason": "Gemma3TextConfig API discrepancy is unrelated to Mistral-3 image preprocessor/model dtype mismatch." + "reason": "Gemma3TextConfig API inconsistency and Mistral-3 image-preprocessor dtype mismatch are different subsystems." }, { "left": "issue:44610", "right": "issue:46369", "accept": false, - "reason": "OmDet-Turbo image-size mismatch and sam3-litetext memory overflow are different problems." + "reason": "Processor output size mismatch and SAM memory overflow are not the same underlying defect." }, { "left": "issue:43650", "right": "issue:46202", "accept": false, - "reason": "The 'ADD THE DATA' issue is too vague to match Gemma pipeline roundtrip failure; no concrete shared bug." + "reason": "A placeholder/low-signal issue and a Gemma pipeline roundtrip bug do not describe the same problem." }, { "left": "issue:42175", "right": "issue:43316", "accept": false, - "reason": "Backend packaging/import issue for tensorflow_text is unrelated to the Gemma3TextConfig API discrepancy." + "reason": "Packaging/backend omission and config API discrepancy are unrelated." }, { "left": "issue:43295", "right": "issue:44871", "accept": false, - "reason": "Processor/tokenizer regression and Gemma eos_token_id inconsistency are different multimodal/tokenization issues." + "reason": "These both involve tokenizer/processor behavior, but one is a regression in processor.tokenizer/image passing and the other is an eos_token_id config inconsistency." }, { "left": "issue:45803", "right": "issue:46365", "accept": false, - "reason": "Bare except swallowing interrupts in FuyuBatchFeature is unrelated to Trainer import failure involving datasets." + "reason": "Bare except swallowing interrupts and import failure until uninstalling datasets are different bugs." }, { "left": "issue:43540", "right": "issue:44514", "accept": false, - "reason": "Qwen3OmniMoe video-input ValueError and Qwen2_5_VLProcessor batched apply_chat_template crash are distinct processor bugs." + "reason": "Qwen3OmniMoe video-processing error and batched apply_chat_template padding crash are separate code paths." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure after align_special_tokens are different model/tokenization bugs." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage and a janus image-generation test failure are unrelated model-specific issues." } ] }, @@ -10358,9 +10300,10 @@ "issue:43673", "issue:43761", "issue:43827", + "issue:43931", "issue:43994", + "issue:44060", "issue:44077", - "issue:44162", "issue:44190", "issue:44242", "issue:44479", @@ -10392,8 +10335,6 @@ "issue:46421" ], "soft_pairs": [ - "issue:44871|issue:45584", - "issue:44162|issue:44792", "issue:45803|issue:46170", "issue:41950|issue:43475", "issue:44871|issue:46050", @@ -10420,216 +10361,211 @@ "issue:44479|issue:44734", "issue:43827|issue:43994", "issue:43065|issue:44610", - "issue:41720|issue:45245" + "issue:41720|issue:45245", + "issue:43931|issue:44060" ], - "prepared_review_unit_hash": "3cff294164186a5b1b58ec0a25c31d4871a17774a1e5163418611f9fc6e2a607", + "prepared_review_unit_hash": "102552a2ef134deae8efa8d49fb6a7254ae5c03f826a4cfade35fcad80d6ddea", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 36019, - "estimated_input_tokens": 9005, - "estimated_eval_tokens": 18266 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36344, + "estimated_input_tokens": 9086, + "estimated_eval_tokens": 18428 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Most items are unrelated issues that only share broad Transformers/v5 vocabulary. The only plausible duplicate pair is the chunked-prefill cache regression cluster around issues 43673 and 46421.", - "confidence": 0.66, - "canonical_issue_reason": "issue:46421 is the best representative because it describes the broader root cause (lazy cache initialization during chunked prefill) and likely subsumes the narrower symptom reported in issue:43673.", + "summary": "Mostly unrelated issue set across generation, model loading, video, config, and backend regressions. The only strong duplicate-looking pair is the chunked-prefill cache bug (43673/46421).", + "confidence": 0.73, + "canonical_issue_reason": "issue:46421 is the best representative of the only clear duplicate pair: it uses the broader, more current wording around cache lazy initialization and chunked prefill.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46421 is the strongest cluster anchor: it is the most general, clearly scoped cache/chunked-prefill regression and reads like the underlying problem rather than a symptom report.", + "best_issue_reason": "issue:46421 is the most general and current issue among the likely duplicates, and it best captures the shared cache/chunked-prefill problem.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44871", - "right": "issue:45584", - "accept": false, - "reason": "Different models and different failures: Gemma EOS config vs Whisper empty-transcription generation." - }, - { - "left": "issue:44162", - "right": "issue:44792", - "accept": false, - "reason": "ESM2 breakage and a Janus test failure are unrelated code paths." - }, { "left": "issue:45803", "right": "issue:46170", "accept": false, - "reason": "Bare-except tensor conversion bug vs Windows/ROCm import dependency error." + "reason": "Different bugs: exception handling in Fuyu preprocessing vs Windows/ROCm import dependency failure." }, { "left": "issue:41950", "right": "issue:43475", "accept": false, - "reason": "Video processor lookup vs SAM3 video output attribute error; not the same bug." + "reason": "Different pipelines/models: video-classification processor lookup vs SAM3 video missing attribute." }, { "left": "issue:44871", "right": "issue:46050", "accept": false, - "reason": "EOS-token config mismatch and LoRA merge collapse are different problems." + "reason": "Unrelated symptoms: Gemma eos_token_id mismatch vs quantized model save/load corruption." }, { "left": "issue:43404", "right": "issue:43441", "accept": false, - "reason": "Weight tying in Mistral3 is separate from FlashAttention failure in Ministral-3." + "reason": "Different code paths: tied lm_head weights vs FlashAttention failure in Ministral-3." }, { "left": "issue:43646", "right": "issue:44913", "accept": false, - "reason": "Generic v5 model-init breakage and GPTNeoX config reload drift are not the same defect." + "reason": "General v5 custom init breakage is not the same as GPTNeoX rotary_pct not persisting on reload." }, { "left": "issue:43452", "right": "issue:44496", "accept": false, - "reason": "Tokenizer/model loading with gguf_file is different from missing model_type in config." + "reason": "Different loading failures: gguf_file handling vs missing model_type in config.json." }, { "left": "issue:43065", "right": "issue:44617", "accept": false, - "reason": "Dummy Conv2d in SAM3 pixel decoder is unrelated to SAM3Video OOM." + "reason": "SAM3 PixelDecoder dummy layer issue is unrelated to SAM3Video OOM." }, { "left": "issue:45127", "right": "issue:45200", "accept": false, - "reason": "LoRA extended-vocab collapse and Gemma 4 mm_token_type_ids defaulting are different issues." + "reason": "LoRA/vocab tie-word-embeddings bug is unrelated to Gemma 4 mm_token_type_ids defaults." }, { "left": "issue:43761", "right": "issue:44479", "accept": false, - "reason": "CLIPVision hidden_states regression and Qwen video-input regression are separate failures." + "reason": "CLIP hidden_states regression and Qwen video input regression are different model behaviors." }, { "left": "issue:43572", "right": "issue:46178", "accept": false, - "reason": "StableLm pad_token_idx config regression is unrelated to tensorflow_text backend mapping." + "reason": "StableLm pad_token_idx config issue is unrelated to tensorflow_text backend mapping." }, { "left": "issue:43504", "right": "issue:45698", "accept": false, - "reason": "Legacy BEIT preset loading and wrong custom module loading are different save/load bugs." + "reason": "Legacy field loading in BEiT is different from wrong custom module resolution after save/load." }, { "left": "issue:45542", "right": "issue:46234", "accept": false, - "reason": "TensorBoard-without-TensorFlow backend error is unrelated to missing import_utils helper." + "reason": "TensorBoard/TensorFlow backend detection bug is unrelated to missing import_utils helper." }, { "left": "issue:43232", "right": "issue:43388", "accept": false, - "reason": "Generation sync_gpus kwargs handling and metric label gathering are different subsystems." + "reason": "Generation kwargs syncing bug is unrelated to gather_for_metrics dropping tuple label elements." }, { "left": "issue:44077", "right": "issue:44534", "accept": false, - "reason": "patchtsmixer post_init validation is not the same as non-persistent buffer initialization junk." + "reason": "Optional post_init enforcement is unrelated to non-persistent buffer initialization corruption." }, { "left": "issue:44792", "right": "issue:44977", "accept": false, - "reason": "Janus image-generation test failure and Qwen3.5 flash-attention generation bug are distinct." + "reason": "Janus test failure and Qwen3.5 FlashAttention generation issue are different bugs." }, { "left": "issue:43761", "right": "issue:44242", "accept": false, - "reason": "Vision hidden_states regression does not match MoE load-balancing loss omission." + "reason": "CLIP vision hidden_states regression is unrelated to router logits/load-balancing loss handling." }, { "left": "issue:44792", "right": "issue:45600", "accept": false, - "reason": "Janus test failure and Sam3LiteText auto_mappings CI break are unrelated." + "reason": "A Janus model test failure is not the same as stale auto_mappings references in CI." }, { "left": "issue:43673", "right": "issue:46421", "accept": true, - "reason": "Both point to the same chunked-prefill cache initialization regression; 46421 states the broader root cause." + "reason": "Both describe the same chunked-prefill cache initialization problem in generation; one frames it as missing cache, the other as lazy init causing recompilation." }, { "left": "issue:44190", "right": "issue:44964", "accept": false, - "reason": "Local dataset loading in a training script is unrelated to loading a multimodal model." + "reason": "Local dataset loading in a training script is unrelated to Phi-4 multimodal model loading." }, { "left": "issue:44617", "right": "issue:45584", "accept": false, - "reason": "SAM3Video OOM and Whisper empty-transcription generation failure have no shared code path." + "reason": "SAM3Video OOM is unrelated to Whisper empty transcription after align_special_tokens." }, { "left": "issue:43673", "right": "issue:44743", "accept": false, - "reason": "Chunked-prefill cache missing and Qwen3.5 recurrent-state reset are different cache bugs." + "reason": "Chunked-prefill cache initialization is not the same as recurrent states resetting with seq_len>1." }, { "left": "issue:44534", "right": "issue:44913", "accept": false, - "reason": "Non-persistent buffer junk and rotary_pct reload defaults are separate config/state issues." + "reason": "Non-persistent buffer junk in v5 is unrelated to GPTNeoXConfig rotary_pct reload behavior." }, { "left": "issue:44855", "right": "issue:45588", "accept": false, - "reason": "Python 3.13 JIT parsing/indentation error is unrelated to flash_attention s_aux handling." + "reason": "Python 3.13 torch.jit parsing error is unrelated to flash_attention s_aux=None crash." }, { "left": "issue:44479", "right": "issue:44734", "accept": false, - "reason": "Video-input regression for Qwen VL models and KV-cache continuation indexing crash are different bugs." + "reason": "Video input regression in Qwen models is unrelated to KV-cache continuation tensor indexing in serve." }, { "left": "issue:43827", "right": "issue:43994", "accept": false, - "reason": "Docs still mentioning pipeline() is not the same as SigLIP2 pipeline/model output corruption." + "reason": "Docs still referencing pipeline() is unrelated to SIGLIP2 producing nonsensical outputs." }, { "left": "issue:43065", "right": "issue:44610", "accept": false, - "reason": "SAM3 pixel-decoder dummy layer and OmDet-Turbo processor size mismatch are unrelated." + "reason": "SAM3 PixelDecoder dummy layer issue is unrelated to OmDet-Turbo processor/model input-size mismatch." }, { "left": "issue:41720", "right": "issue:45245", "accept": false, - "reason": "Qwen3 auto device mapping CUDA assert and category-count overflow are completely different failures." + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to category-count overflow." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch is unrelated to Qwen3-Next tied-weights warning." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet and stays conservative. The 43673/46421 cache/chunked-prefill pair is the only clearly plausible duplicate cluster from the provided titles, and the other soft-edge rejections do not overclaim equivalence. No unsupported mergeability claims appear." + "feedback": "Grounded and conservative overall. The summary matches the packet, and the only accepted soft pair (43673/46421) is reasonably supported by the titles as the same cache/chunked-prefill bug. The other rejections are appropriately cautious and not overstated." }, "evaluator_used": true, "retried": false, @@ -10663,12 +10599,11 @@ "issue:43540", "issue:43550", "issue:43644", - "issue:43931", "issue:43976", - "issue:44060", "issue:44077", "issue:44112", "issue:44190", + "issue:44246", "issue:44462", "issue:44521", "issue:44617", @@ -10679,6 +10614,7 @@ "issue:44877", "issue:44913", "issue:44938", + "issue:45003", "issue:45127", "issue:45216", "issue:45245", @@ -10699,7 +10635,6 @@ "issue:46489" ], "soft_pairs": [ - "issue:43931|issue:44060", "issue:45381|issue:46369", "issue:35141|issue:42490", "issue:43976|issue:44855", @@ -10725,204 +10660,205 @@ "issue:36246|issue:41628", "issue:46082|issue:46121", "issue:43404|issue:45216", - "issue:45245|issue:45584" + "issue:45245|issue:45584", + "issue:44246|issue:45003" ], - "prepared_review_unit_hash": "5f4ea19d0fe200cdbe764101bd84a43fe74f26a0d982846b96cd0b80a6f22311", + "prepared_review_unit_hash": "c5203eebd6489b574618466d4f1e99c7e6775a8283ba812748266cfe28d5d01c", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 27, - "serialized_chars": 35519, - "estimated_input_tokens": 8880, - "estimated_eval_tokens": 18016 + "serialized_chars": 35423, + "estimated_input_tokens": 8856, + "estimated_eval_tokens": 17968 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly unrelated issue reports with only one plausible duplicate pair: the DeepSeek tokenizer regression in v5. The rest share superficial keywords (tokenizer, config, model loading, memory, Python versions) but describe different models or code paths.", - "confidence": 0.92, - "canonical_issue_reason": "issue:46489 is the most specific and actionable DeepSeek tokenizer regression report: it names the affected model family, the v5 breakage, and the remaining gap in PR #44801's fix.", + "summary": "Most pairs are only loosely related by subsystem or error type and should not be deduplicated. The only strong duplicate candidate is the DeepSeek tokenizer regression pair: issue 44779 and issue 46489 describe the same underlying v5 tokenizer breakage, with 46489 reading like a more specific follow-up to the broader 44779 report.", + "confidence": 0.87, + "canonical_issue_reason": "Issue 44779 is the best canonical because it is the broader, earlier DeepSeek tokenizer regression report; 46489 appears to be a narrower reproduction/follow-up for the same v5 incorrect-output bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:46489 is the best representative issue for the only real duplicate cluster because it pinpoints the concrete v5 tokenizer regression and the unresolved fix gap.", + "best_issue_reason": "44779 is the strongest issue to keep as the umbrella duplicate target for the DeepSeek tokenizer regression cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43931", - "right": "issue:44060", - "accept": false, - "reason": "Both mention Qwen and weights, but they are different model families and different bugs: a load-time shape mismatch vs an incorrect tied-weights warning." - }, { "left": "issue:45381", "right": "issue:46369", "accept": false, - "reason": "Different subsystems and symptoms: Qwen2.5-VL video position IDs vs Sam3-LiteText memory overflow." + "reason": "Both are multimodal/video-related, but one is a wrong vision-position-id bug and the other is a memory overflow/OOM report; not the same defect." }, { "left": "issue:35141", "right": "issue:42490", "accept": false, - "reason": "Unrelated bugs: embedding reinitialization after resize vs import warning for trailing-slash local paths." + "reason": "These are unrelated: one is about output embeddings being reinitialized on resize, the other about a local-model path ending in a slash triggering an import warning." }, { "left": "issue:43976", "right": "issue:44855", "accept": false, - "reason": "Both involve Python compatibility, but one is a version support issue and the other is a specific IndentationError in a scripted model import." + "reason": "Both mention Python-version compatibility, but they are different failure modes in different code paths: package loading on Python 3.9+ vs an IndentationError in DebertaV2 import on Python 3.13." }, { "left": "issue:43644", "right": "issue:44913", "accept": false, - "reason": "Different problems: non-persistent buffer initialization in Transformers 5.0 vs GPTNeoXConfig losing rotary_pct on reload." + "reason": "One is about non-persistent buffers being filled with junk in Transformers 5.0.0; the other is about a config value not persisting on reload. Similar theme, different bug." }, { "left": "issue:43550", "right": "issue:44190", "accept": false, - "reason": "Completely different code paths: torch.compile/SDPA failure in Bamba vs local dataset loading in an example script." + "reason": "These are unrelated runtime/loading failures in different areas: Bamba SDPA torch.compile vs loading a local dataset in an example script." }, { "left": "issue:44077", "right": "issue:44877", "accept": false, - "reason": "Both concern config handling, but they are different model-specific loading issues and not the same underlying bug." + "reason": "Both touch config validation, but the actual problems differ: optional post_init handling for patchtsmixer vs strict config blocking granite_speech loading." }, { "left": "issue:45127", "right": "issue:45357", "accept": false, - "reason": "Different save/merge paths and different models: LoRA merge collapse vs incorrect visual encoder keys in save_pretrained." + "reason": "Both involve save/load or embedding-related regressions, but they affect different models and code paths; not mergeable as one concrete fix." }, { "left": "issue:44617", "right": "issue:46369", "accept": false, - "reason": "Same broad Sam3 family, but OOM in Sam3Video is not the same bug as memory overflow in Sam3-LiteText." + "reason": "Both are memory-related, but one is a specific CUDA OOM for Sam3Video and the other is a memory overflow in sam3-litetext; not the same bug." }, { "left": "issue:43540", "right": "issue:44734", "accept": false, - "reason": "Unrelated: Qwen3OmniMoe video input validation failure vs serving KV-cache continuation indexing bug." + "reason": "Both involve video/KV-cache handling at a high level, but they describe different crashes in different model/server paths." }, { "left": "issue:25251", "right": "issue:27957", "accept": false, - "reason": "No common underlying bug: pipeline top_k nesting change vs Flash Attention 2 support for XLMRoberta." + "reason": "A pipeline output-shape regression and Flash Attention 2 support for XLMRoberta are unrelated." }, { "left": "issue:44779", "right": "issue:46489", "accept": true, - "reason": "Same underlying DeepSeek tokenizer regression in Transformers v5; 46489 is a more specific form of the same broken tokenizer behavior and references the partial fix gap." + "reason": "Both report DeepSeek tokenizer producing incorrect output in v5; 46489 explicitly frames the issue as a gap in the earlier fix, so they look like the same underlying regression." }, { "left": "issue:43335", "right": "issue:44190", "accept": false, - "reason": "Different bugs: SwitchTransformers sparse-layer creation on zero sparse layers vs dataset loading failure in an image-classification script." + "reason": "A SwitchTransformers sparse-layer construction bug and a local dataset loading issue are unrelated." }, { "left": "issue:44112", "right": "issue:45600", "accept": false, - "reason": "Both touch CI/test breakage, but the underlying causes are different and model-specific." + "reason": "Both mention CI, but one is a stale test failure in GraniteSpeech and the other is auto_mappings referencing removed configs; different artifacts and fixes." }, { "left": "issue:45632", "right": "issue:46234", "accept": false, - "reason": "Different API issues: cache-path collision for trust_remote_code vs a missing availability helper in import_utils." + "reason": "A cache-path collision for trust_remote_code and a missing import utility function are unrelated." }, { "left": "issue:44462", "right": "issue:45907", "accept": false, - "reason": "Both involve Hub/file listing behavior, but one is AutoTokenizer ignoring tokenizer.json and the other is list_repo_files missing revision forwarding." + "reason": "Tokenizer repo resolution and list_repo_files revision forwarding are different hub/API issues." }, { "left": "issue:44737", "right": "issue:45803", "accept": false, - "reason": "Different bugs: XLNet device placement in relative positional encoding vs bare except swallowing interrupts/errors in FuyuBatchFeature." + "reason": "XLNet CPU allocation and a bare-except bug in Fuyu tensor conversion are unrelated." }, { "left": "issue:43116", "right": "issue:44190", "accept": false, - "reason": "Different example scripts and failure modes: empty multi-label classification results vs local dataset loading failure." + "reason": "One is multi-label classification returning empty results, the other is a dataset-loading problem; not the same bug." }, { "left": "issue:44938", "right": "issue:45542", "accept": false, - "reason": "Unrelated environment/backend issues: Python 3.14 import failure vs tensorboard-only install causing a TensorFlow backend error." + "reason": "Python 3.14 import/load failure and an undefined tf backend error from tensorboard-only installs are different compatibility issues." }, { "left": "issue:44521", "right": "issue:45381", "accept": false, - "reason": "Both are multimodal issues, but the bug details differ: assistant masks in chat templates vs Qwen2.5-VL video position IDs." + "reason": "Both are multimodal-template related, but one is all-zero assistant masks and the other is wrong video position IDs; different outputs and code paths." }, { "left": "issue:43295", "right": "issue:45276", "accept": false, - "reason": "Different regression types and models: processor/tokenizer API breakage vs Gemma4 resize_token_embeddings not updating per-layer/output embeddings." + "reason": "A regression around processor.tokenizer/image passing and a gemma4 resize-token-embeddings bug are not the same defect." }, { "left": "issue:45276", "right": "issue:45584", "accept": false, - "reason": "Both involve generation/embeddings behavior, but the reported failures are unrelated." + "reason": "These are unrelated: embedding resize propagation vs Whisper generation failing on empty transcription." }, { "left": "issue:45632", "right": "issue:46421", "accept": false, - "reason": "Different cache/path behaviors: trust_remote_code cache collision vs lazy cache initialization causing recompilation." + "reason": "A trust_remote_code cache-path collision and spurious recompilation from cache lazy init are separate issues." }, { "left": "issue:36246", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but for different symbols and different submodules; not the same missing export bug." + "reason": "Both are import errors, but they involve different missing symbols and different module surfaces." }, { "left": "issue:46082", "right": "issue:46121", "accept": false, - "reason": "Different validation issues: LlamaConfig head_dim constraints vs rope validation rejecting JSON-loaded lists." + "reason": "LlamaConfig head_dim validation and rope-parameter conversion on JSON-loaded lists are different configuration bugs." }, { "left": "issue:43404", "right": "issue:45216", "accept": false, - "reason": "Different save/tie-weight bugs in different model families; not the same underlying issue." + "reason": "Both involve model save/tie-weight behavior, but one is missing lm_head tying in Mistral3 and the other is incorrect Qwen3.5 save_pretrained output; not the same bug." }, { "left": "issue:45245", "right": "issue:45584", "accept": false, - "reason": "No shared bug: category cardinality limit vs Whisper empty-transcription generation failure." + "reason": "Category-count runtime limit and Whisper empty-transcription generation failure are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Slow import and unsafe sys.modules access are different import-time problems with different fixes." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and stays conservative on the soft-edge pairs. The single accepted duplicate pair, issue:44779 and issue:46489, is supported by the titles as the same DeepSeek tokenizer regression in v5, and the rest of the verdicts avoid overstating similarity." + "feedback": "The summary is well-grounded in the packet and stays conservative on the soft-edge pairs. The only accepted duplicate candidate, issue 44779 vs 46489, is supported by the titles and the follow-up framing in 46489. No soft-edge acceptance appears overstated, and the non-duplicates are rejected for plausible, specific reasons." }, "evaluator_used": true, "retried": false, @@ -10954,6 +10890,7 @@ "issue:43295", "issue:43299", "issue:43381", + "issue:43388", "issue:43441", "issue:43479", "issue:43526", @@ -10961,13 +10898,13 @@ "issue:43643", "issue:43650", "issue:43673", + "issue:43688", "issue:43854", "issue:43901", "issue:44016", "issue:44112", "issue:44186", "issue:44220", - "issue:44246", "issue:44279", "issue:44336", "issue:44464", @@ -10978,7 +10915,6 @@ "issue:44792", "issue:44933", "issue:44964", - "issue:45003", "issue:45042", "issue:45216", "issue:45335", @@ -10993,7 +10929,6 @@ "issue:46424" ], "soft_pairs": [ - "issue:44246|issue:45003", "issue:44933|issue:45362", "issue:44186|issue:44734", "issue:44683|issue:45446", @@ -11021,210 +10956,232 @@ "issue:42175|issue:44279", "issue:41950|issue:44016", "issue:43381|issue:45593", - "issue:43901|issue:44496" + "issue:43901|issue:44496", + "issue:43388|issue:43688", + "issue:42796|issue:43650", + "issue:44016|issue:44220", + "issue:45335|issue:45584" ], - "prepared_review_unit_hash": "21c97788ce0148726f79aef948f4af0626534c80d0787951c6511fbd9727ee93", + "prepared_review_unit_hash": "96321b1e6e1e48dcec00089a6737592e6279d78b1e85d9523167f7af71309c16", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36222, - "estimated_input_tokens": 9056, - "estimated_eval_tokens": 18368 + "soft_pair_count": 32, + "serialized_chars": 37225, + "estimated_input_tokens": 9307, + "estimated_eval_tokens": 18870 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This set is a heterogeneous grab-bag of unrelated closed issues. None of the proposed soft pairs look like the same underlying bug or a single mergeable fix, so all soft edges should be rejected.", - "confidence": 0.97, - "canonical_issue_reason": "No single canonical issue stands out because the items span many unrelated subsystems (imports, config loading, tokenization, generation, cache handling, pipelines, and docs).", + "summary": "This cluster is a grab bag of unrelated issue reports, mostly around model loading, generation, configs, and a few packaging/import regressions. I do not see any soft pair that clearly describes the same underlying bug or change, so all candidate duplicate edges should be rejected.", + "confidence": 0.82, + "canonical_issue_reason": "44964 is the clearest, most general user-facing bug report in the set: a concrete model-loading regression against the latest transformers release. It is well-scoped and representative of the broader class of breakages here.", "canonical_pr_reason": null, - "best_issue_reason": "No best issue candidate: nothing here clearly subsumes the others or represents the same bug family.", + "best_issue_reason": "44964 is the best standalone issue because it is specific, reproducible, and broadly relevant to users hitting model loading failures after upgrading.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44246", - "right": "issue:45003", - "accept": false, - "reason": "Both relate to imports/performance, but one is about slow import timing and the other about unsafe sys.modules access; not the same bug." - }, { "left": "issue:44933", "right": "issue:45362", "accept": false, - "reason": "Different failures in different areas: a nonexistent image_utils import vs Qwen3.5 chat model crashes." + "reason": "Different failures: missing image_utils import vs Qwen chat/runtime crash. No shared code path." }, { "left": "issue:44186", "right": "issue:44734", "accept": false, - "reason": "Tokenizer/NER padding crash is unrelated to a serve KV-cache continuation indexing bug." + "reason": "Tokenizer padding/NER crash is unrelated to serve KV-cache continuation indexing." }, { "left": "issue:44683", "right": "issue:45446", "accept": false, - "reason": "Both touch flex_attention, but one is a torch>=2.9 compile/runtime failure and the other is a version-check/import-path bug." + "reason": "Both touch flex_attention, but one is a torch>=2.9 compile failure and the other is a version-check/import bug. Different concrete problems." }, { "left": "issue:43673", "right": "issue:46424", "accept": false, - "reason": "Both mention cache, but one is missing GenerationMixin cache during chunked prefill and the other is max_cache_length handling for quantized cache." + "reason": "Both mention cache, but one is missing GenerationMixin cache in chunked prefill and the other is max_cache_length only applying to quantized cache." }, { "left": "issue:44464", "right": "issue:45563", "accept": false, - "reason": "Chunked-generation output inconsistency is not the same as a stale warning about num_return_sequences in paged generate()." + "reason": "Chunked generation inconsistency with compiled forward is unrelated to a stale warning about num_return_sequences." }, { "left": "issue:44964", "right": "issue:45600", "accept": false, - "reason": "Model loading for Phi-4-multimodal is unrelated to auto_mappings referencing removed Sam3LiteText configs." + "reason": "Different model/config regressions: Phi-4 multimodal loading vs removed Sam3LiteText auto-mapping references." }, { "left": "issue:43854", "right": "issue:44964", "accept": false, - "reason": "Both are model-loading issues, but they involve different model families and different failure modes." + "reason": "Both are model-loading issues, but for different model families and likely different failures." }, { "left": "issue:43643", "right": "issue:44496", "accept": false, - "reason": "One is about trust_remote_code dropping fields; the other is an unrecognized model_type in config.json." + "reason": "AutoConfig trust_remote_code field loss is unrelated to unrecognized OLMo-Hybrid config loading." }, { "left": "issue:44112", "right": "issue:44964", "accept": false, - "reason": "CI stale-device override failure in GraniteSpeech is unrelated to Phi-4-multimodal loading." + "reason": "CI device override test failure is not the same as a model-load regression." }, { "left": "issue:43650", "right": "issue:44220", "accept": false, - "reason": "Placeholder text issue is unrelated to _torch_extract_fbank_features() behavior." + "reason": "No meaningful overlap: generic 'ADD THE DATA' issue vs fbank feature extraction bug." }, { "left": "issue:36246", "right": "issue:38617", "accept": false, - "reason": "Both are ImportErrors, but for different missing symbols in unrelated modules." + "reason": "Separate import errors in different modules with no shared code path." }, { "left": "issue:42796", "right": "issue:44521", "accept": false, - "reason": "Tokenizer BPE-marker handling is not the same as multimodal assistant_mask generation in apply_chat_template()." + "reason": "Tokenizer BPE marker handling and multimodal assistant mask generation are different bugs." }, { "left": "issue:27957", "right": "issue:35141", "accept": false, - "reason": "Flash Attention support for XLMRoberta is unrelated to embedding resize reinitialization in post_init." + "reason": "Flash Attention 2 support for XLM-R is unrelated to embedding reinitialization after resize_token_embeddings." }, { "left": "issue:43479", "right": "issue:45375", "accept": false, - "reason": "Both involve config defaults, but one is None-handling for Phi4Multimodal and the other is strict field dropping in Qwen3_5MoeVisionConfig." + "reason": "Different config-field issues in different model families; not the same bug." }, { "left": "issue:45563", "right": "issue:45584", "accept": false, - "reason": "A stale warning in paged generate is unrelated to Whisper failing on empty transcription after align_special_tokens." + "reason": "A stale paging warning is unrelated to Whisper empty-transcription generation failure." }, { "left": "issue:43531", "right": "issue:45600", "accept": false, - "reason": "Sliding_window behavior for Qwen3-MoE is unrelated to removed Sam3LiteText config references." + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup are unrelated." }, { "left": "issue:43295", "right": "issue:45335", "accept": false, - "reason": "Custom processor/tokenizer regression is unrelated to t5gemma resize_token_embeddings not updating decoder.embed_tokens." + "reason": "Custom processor/tokenizer regression is unrelated to t5gemma decoder embedding resize behavior." }, { "left": "issue:43526", "right": "issue:43901", "accept": false, - "reason": "BeitImageProcessorFast reduce_labels bug is unrelated to a docs mismatch for return_all_scores." + "reason": "BeitImageProcessorFast label reduction bug is unrelated to stale documentation wording." }, { "left": "issue:42907", "right": "issue:43441", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals is unrelated to Ministral FlashAttention inference failure." + "reason": "Both involve Ministral/Devstral or flash attention, but one is save/load of dequantized weights and the other is inference failure under FlashAttention; not the same bug." }, { "left": "issue:43441", "right": "issue:45588", "accept": false, - "reason": "Both mention flash attention, but the concrete failures differ: Ministral model loading vs s_aux=None crashing in flash_attention.py." + "reason": "Both mention flash attention, but the reported failures are different: model-specific RC failure vs AttributeError on s_aux=None in sink-less models." }, { "left": "issue:44496", "right": "issue:45042", "accept": false, - "reason": "Unrecognized model_type loading error is unrelated to PIL image processors requiring torchvision." + "reason": "Unrecognized model config and PIL image-processor torchvision dependency are unrelated." }, { "left": "issue:43673", "right": "issue:44336", "accept": false, - "reason": "Generation cache issues are unrelated to ANSI codes emitted by loading_report." + "reason": "Cache handling in generation is unrelated to ANSI escape codes in loading_report." }, { "left": "issue:43299", "right": "issue:45216", "accept": false, - "reason": "Qwen3VL MoE loading breakage is unrelated to Qwen3.5 save_pretrained checkpoint corruption." + "reason": "Qwen3VL loading failure and Qwen3.5 save_pretrained checkpoint corruption are different issues." }, { "left": "issue:44792", "right": "issue:46424", "accept": false, - "reason": "A janus image-generation test failure is unrelated to cache max length behavior." + "reason": "Janus image-generation test failure is unrelated to cache length configuration behavior." }, { "left": "issue:42175", "right": "issue:44279", "accept": false, - "reason": "Missing TensorFlow in the torch extra is not the same as a generic dependency issue." + "reason": "Package-extra backend composition issue and generic dependency issue are too broad and not the same concrete bug." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification pipeline bug is unrelated to a notebook syntax error." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing eval-mode limitation is unrelated to D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Documentation mismatch for return_all_scores is unrelated to an unrecognized model config error." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Dropped labels in gather_for_metrics and auxiliary-loss normalization are different training bugs." }, { - "left": "issue:41950", - "right": "issue:44016", + "left": "issue:42796", + "right": "issue:43650", "accept": false, - "reason": "A pipeline bug is unrelated to a notebook syntax error." + "reason": "Tokenizer BPE-marker handling has nothing to do with the generic 'ADD THE DATA' issue." }, { - "left": "issue:43381", - "right": "issue:45593", + "left": "issue:44016", + "right": "issue:44220", "accept": false, - "reason": "Gradient checkpointing in eval mode and D-FINE auxiliary-loss behavior are different code paths and bugs." + "reason": "Notebook syntax error and _torch_extract_fbank_features() are unrelated." }, { - "left": "issue:43901", - "right": "issue:44496", + "left": "issue:45335", + "right": "issue:45584", "accept": false, - "reason": "Docs mentioning return_all_scores is unrelated to a model config lacking model_type." + "reason": "t5gemma embedding resize behavior is unrelated to Whisper empty-transcription generation after align_special_tokens." } ] }, @@ -11251,15 +11208,14 @@ "issue:38175", "issue:39692", "issue:41950", - "issue:42796", "issue:43295", "issue:43299", - "issue:43388", + "issue:43441", + "issue:43450", "issue:43479", "issue:43525", "issue:43577", "issue:43650", - "issue:43688", "issue:43824", "issue:43854", "issue:43901", @@ -11267,11 +11223,13 @@ "issue:44016", "issue:44062", "issue:44075", - "issue:44220", "issue:44242", "issue:44279", + "issue:44297", "issue:44351", "issue:44368", + "issue:44462", + "issue:44466", "issue:44496", "issue:44617", "issue:44683", @@ -11282,28 +11240,22 @@ "issue:45103", "issue:45200", "issue:45216", + "issue:45276", "issue:45278", - "issue:45335", "issue:45357", "issue:45372", "issue:45405", "issue:45542", - "issue:45584", "issue:45759", "issue:45812", "issue:45823", + "issue:45834", "issue:46082", "issue:46209", - "issue:46234", "issue:46399" ], "soft_pairs": [ - "issue:43388|issue:43688", - "issue:42796|issue:43650", - "issue:44016|issue:44220", - "issue:45335|issue:45584", "issue:44496|issue:45759", - "issue:45372|issue:46234", "issue:44075|issue:44368", "issue:43577|issue:45216", "issue:46082|issue:46399", @@ -11326,192 +11278,190 @@ "issue:43976|issue:44279", "issue:43299|issue:43854", "issue:43479|issue:43525", - "issue:43976|issue:44987" + "issue:43976|issue:44987", + "issue:43450|issue:44279", + "issue:44297|issue:44462", + "issue:44466|issue:45276", + "issue:43441|issue:45834" ], - "prepared_review_unit_hash": "8eb38b2aa41555e76089cd07beb29fbfb55114e1b8242dc1128d0361a5b49c0b", + "prepared_review_unit_hash": "1afc254e5954e0f6b24758bf9e97412de6826da9b1f0bdc61b9f9292eb144e90", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36442, - "estimated_input_tokens": 9111, - "estimated_eval_tokens": 18478 + "soft_pair_count": 28, + "serialized_chars": 36162, + "estimated_input_tokens": 9041, + "estimated_eval_tokens": 18338 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The set is dominated by distinct, unrelated issues; none of the soft pairs appears to be the same underlying bug or change. The only plausible umbrella is the broad upgrade/import-regression report, but it still does not duplicate the other reports.", - "confidence": 0.72, - "canonical_issue_reason": "issue:45278 is the broadest umbrella for upgrade-induced import failures; the other items are narrower, model- or feature-specific bugs and do not collapse into one root cause.", + "summary": "These are mostly independent Transformers issues spanning different models, configs, pipelines, and regression types. The soft-similar pairs share broad themes (loading, tokenizers, multimodal configs, version regressions) but not the same concrete bug, so none should be merged as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45020 is the broadest umbrella regression report in the set and the best cluster anchor, though it does not actually duplicate the other issues.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 is the best single representative for the cluster because it captures the most general import-breakage symptom across versions, even though it is not a true duplicate of the others.", + "best_issue_reason": "issue:45020 best represents the cluster\u2019s recurring theme of recent-version model-loading regressions.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43388", - "right": "issue:43688", - "accept": false, - "reason": "Different bugs: last-batch label truncation in gather_for_metrics vs auxiliary-loss normalization in OLMoE/GPT OSS." - }, - { - "left": "issue:42796", - "right": "issue:43650", - "accept": false, - "reason": "Unrelated: tokenizer BPE marker handling vs a vague request titled \"ADD THE DATA\"." - }, - { - "left": "issue:44016", - "right": "issue:44220", - "accept": false, - "reason": "Different code paths: notebook syntax error vs audio feature extraction failure." - }, - { - "left": "issue:45335", - "right": "issue:45584", - "accept": false, - "reason": "Different models and failures: t5gemma embedding resize bug vs Whisper empty-transcription generation bug." - }, { "left": "issue:44496", "right": "issue:45759", "accept": false, - "reason": "Both are model-loading/config issues, but one is unrecognized model config and the other is composite multimodal config unwrapping; not the same bug." - }, - { - "left": "issue:45372", - "right": "issue:46234", - "accept": false, - "reason": "Both are import-related, but they involve different missing symbols in different modules and different downstream failures." + "reason": "Both are model-loading/configuration failures, but one is an unrecognized model_type error and the other is a composite-config unwrap bug for Qwen multimodal configs." }, { "left": "issue:44075", "right": "issue:44368", "accept": false, - "reason": "Different symptoms and fixes: SGD args ignored vs a warning about tie_word_embeddings in Qwen fine-tuning." + "reason": "Unrelated problems: optimizer SGD arguments being ignored vs a tie_word_embeddings warning during Qwen fine-tuning." }, { "left": "issue:43577", "right": "issue:45216", "accept": false, - "reason": "Different regressions: dtype propagation in BLIP-2 vs incorrect save_pretrained output for Qwen3.5." + "reason": "Different models and different bugs: BLIP2 dtype propagation on load vs incorrect Qwen3.5 checkpoint serialization." }, { "left": "issue:46082", "right": "issue:46399", "accept": false, - "reason": "Different issues: LlamaConfig validation on head_dim vs Qwen3.5 rotary-kernel registration failure." + "reason": "A config-validation issue for Llama vs a Qwen3.5 kernel registration crash; not the same code path." }, { "left": "issue:45103", "right": "issue:45372", "accept": false, - "reason": "Docstring generation crash with future annotations is unrelated to the mistral_common ReasoningEffort import error." + "reason": "auto_docstring annotation handling crash vs a mistral_common import error affecting Gemma 4 processor loading." }, { "left": "issue:43824", "right": "issue:44933", "accept": false, - "reason": "Both are import errors, but they are for different missing exports and different model areas; not the same root cause." + "reason": "Both involve imports, but they are different missing symbols in different modules and not the same underlying export bug." }, { "left": "issue:44279", "right": "issue:45405", "accept": false, - "reason": "Generic dependency complaint vs a specific unreleased minimum PEFT version bump; not the same bug." + "reason": "A generic dependency issue is not the same as PEFT version pinning to an unreleased release." }, { "left": "issue:44242", "right": "issue:45200", "accept": false, - "reason": "Different training bugs: missing load-balancing loss vs mm_token_type_ids default handling." + "reason": "One is about router load-balancing loss, the other about defaulting mm_token_type_ids for text-only fine-tuning." }, { "left": "issue:38175", "right": "issue:44617", "accept": false, - "reason": "Completely different failures: zero probabilities in SigLIP2 vs CUDA OOM in Sam3Video." + "reason": "Unrelated symptoms: zero probabilities in SigLIP2 vs CUDA OOM in Sam3Video." }, { "left": "issue:41950", "right": "issue:45020", "accept": false, - "reason": "Different subsystems and failures: video-classification processor lookup vs remote_code loading regressions." + "reason": "Video-classification looking for image processors is a pipeline routing bug, not the same as remote_code model loading regressions." }, { "left": "issue:43299", "right": "issue:45357", "accept": false, - "reason": "Both involve Qwen3.5 multimodal models, but one is loading failure and the other is save_pretrained key serialization; not the same bug." + "reason": "Both concern Qwen multimodal models, but one is a loading failure and the other is incorrect save_pretrained key serialization." }, { "left": "issue:44351", "right": "issue:45278", "accept": false, - "reason": "HybridCache import missing is one specific import failure; the other is a broad report of many import errors after upgrade, not the same underlying issue." + "reason": "The specific HybridCache import error may be one symptom among many, but 45278 is a broad import-regression report rather than the same concrete bug." }, { "left": "issue:44016", "right": "issue:44062", "accept": false, - "reason": "Notebook syntax error vs AddedToken argument duplication; unrelated." + "reason": "Notebook syntax error and AddedToken constructor argument collision are unrelated." }, { "left": "issue:43650", "right": "issue:46209", "accept": false, - "reason": "No relation: a vague data request vs FP8 inference failure on B200." + "reason": "Placeholder issue vs FP8 inference failure on B200; no shared bug." }, { "left": "issue:41950", "right": "issue:45823", "accept": false, - "reason": "Different bugs in different areas: video pipeline processor lookup vs Gemma4 device-map mismatch." + "reason": "Video pipeline processor lookup and Gemma4 device placement during forward are distinct failures." }, { "left": "issue:39692", "right": "issue:43901", "accept": false, - "reason": "Both are documentation-related, but they concern different examples and different behaviors." + "reason": "Both are docs-related, but they describe different examples and different API mismatches." }, { "left": "issue:44683", "right": "issue:45542", "accept": false, - "reason": "Unrelated runtime issues: compiled flex_attention on torch 2.9 vs tensorboard/TensorFlow backend mismatch." + "reason": "Compiled flex_attention on new torch versions is unrelated to a TensorFlow/backend dependency problem." }, { "left": "issue:43295", "right": "issue:44779", "accept": false, - "reason": "Both are tokenizer regressions, but they affect different paths/models and do not look like one fixable bug." + "reason": "Custom processor.tokenizer access regression and Deepseek tokenizer output regression are different code paths and different models." }, { "left": "issue:43976", "right": "issue:45405", "accept": false, - "reason": "Version/Python support issue vs minimum PEFT version dependency bump; unrelated." + "reason": "Python version support failure is not the same as a PEFT minimum-version bump." }, { "left": "issue:44779", "right": "issue:45812", "accept": false, - "reason": "Different tokenizer regressions on different model families; too broad to treat as one duplicate." + "reason": "Both are tokenizer regressions, but for different model families and different failure modes; not one duplicate bug." }, { "left": "issue:43976", "right": "issue:44279", "accept": false, - "reason": "Python version support failure vs a generic dependency issue; unrelated." + "reason": "A runtime Python compatibility issue and a dependency issue are not the same underlying problem." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Both are model-loading issues, but they affect different models and have different root causes." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Default config initialization for Phi4Multimodal is unrelated to the Llama4Config missing pad_token_id attribute." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Different bugs around embeddings/serialization: device-dependent lm_head.weight saving vs resize_token_embeddings not updating Gemma4 layer/output embeddings." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "FlashAttention failure in Ministral-3 and long-input OCR index error in Kosmos2.5 are unrelated." } ] }, @@ -11540,11 +11490,13 @@ "issue:42994", "issue:43317", "issue:43441", - "issue:43450", "issue:43525", "issue:43643", "issue:43653", + "issue:43673", "issue:43723", + "issue:43761", + "issue:43784", "issue:43881", "issue:43901", "issue:43976", @@ -11555,13 +11507,10 @@ "issue:44246", "issue:44261", "issue:44263", - "issue:44279", "issue:44291", "issue:44297", "issue:44336", "issue:44361", - "issue:44462", - "issue:44466", "issue:44521", "issue:44589", "issue:44623", @@ -11570,25 +11519,22 @@ "issue:44749", "issue:44779", "issue:45081", - "issue:45276", "issue:45325", "issue:45362", "issue:45397", + "issue:45542", "issue:45593", "issue:45600", "issue:45656", "issue:45712", "issue:45823", - "issue:45834", "issue:45907", "issue:46123", - "issue:46369" + "issue:46291", + "issue:46369", + "issue:46399" ], "soft_pairs": [ - "issue:43450|issue:44279", - "issue:44297|issue:44462", - "issue:44466|issue:45276", - "issue:43441|issue:45834", "issue:45081|issue:45907", "issue:43653|issue:45593", "issue:44623|issue:44749", @@ -11605,6 +11551,7 @@ "issue:44671|issue:44779", "issue:43901|issue:43994", "issue:43723|issue:44162", + "issue:46291|issue:46399", "issue:44297|issue:45397", "issue:43525|issue:43881", "issue:43901|issue:44246", @@ -11612,204 +11559,200 @@ "issue:43525|issue:43643", "issue:44589|issue:45362", "issue:44188|issue:44336", - "issue:41628|issue:43976" + "issue:41628|issue:43976", + "issue:43784|issue:45542", + "issue:43673|issue:43761" ], - "prepared_review_unit_hash": "0a1aa968050bf2c83e7c12b736f4eec324aeb9b597e172121c51b2460c9f4fc9", + "prepared_review_unit_hash": "491902b07895feabb0afced74ca0e8093d3ff17ca9ead58db73d826a155af929", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 28, - "serialized_chars": 35941, - "estimated_input_tokens": 8986, - "estimated_eval_tokens": 18228 + "soft_pair_count": 27, + "serialized_chars": 35786, + "estimated_input_tokens": 8947, + "estimated_eval_tokens": 18150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: most items are unrelated Transformers bugs, docs issues, tokenizer/model regressions, or environment-specific failures. None of the soft pairs look like the same underlying bug/change, so all should be rejected as duplicates.", + "summary": "The set is heterogeneous, and every soft pair appears to be a false positive: they share broad themes like tokenizers, loading, or deepspeed, but not the same concrete bug/change.", "confidence": 0.96, - "canonical_issue_reason": "No true duplicate hub stands out; if one issue must anchor the set, issue:44521 is the most active/reference-linked, but the cluster itself is not coherent enough to justify a real canonical duplicate.", + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "issue:44521 has the strongest discussion/inbound-reference signal and is the most likely to attract follow-up reports, even though it is not a good representation of the rest of the heterogeneous set.", + "best_issue_reason": null, "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43450", - "right": "issue:44279", - "accept": false, - "reason": "Different problems: batched video processor shape vs generic dependency trouble." - }, - { - "left": "issue:44297", - "right": "issue:44462", - "accept": false, - "reason": "Both involve tokenizer saving/loading, but one is tokenizer_class metadata mismatch and the other is AutoTokenizer ignoring tokenizer.json." - }, - { - "left": "issue:44466", - "right": "issue:45276", - "accept": false, - "reason": "Different failures: lm_head serialization inconsistency vs resize_token_embeddings not updating Gemma4 embeddings/output layers." - }, - { - "left": "issue:43441", - "right": "issue:45834", - "accept": false, - "reason": "FlashAttention model load/runtime bug vs Kosmos2.5 OCR long-input index error." - }, { "left": "issue:45081", "right": "issue:45907", "accept": false, - "reason": "Mistral regex patch crash is unrelated to list_repo_files failing to forward revision." + "reason": "Both touch tokenizer/revision handling, but one is a Mistral regex crash during load while the other is a repo_files API bug in CI." }, { "left": "issue:43653", "right": "issue:45593", "accept": false, - "reason": "Tokenizer special-token registration bug vs D-FINE auxiliary-loss behavior; not the same bug." + "reason": "Completely different areas: BigBirdTokenizer special-token decode vs D-FINE auxiliary-loss behavior." }, { "left": "issue:44623", "right": "issue:44749", "accept": false, - "reason": "Processor save_pretrained missing files is unrelated to a Chinese-language performance regression report." + "reason": "One is processor save_pretrained file generation; the other is a performance regression after upgrade." }, { "left": "issue:42994", "right": "issue:43317", "accept": false, - "reason": "Quantized model saving failure vs dequantized model loading/offload failure; related area, different code paths." + "reason": "Both involve quantization/offload, but save failure and load failure are different code paths and symptoms." }, { "left": "issue:45600", "right": "issue:45712", "accept": false, - "reason": "Both are CI/cleanup-like, but one is stale auto_mappings config references and the other is leftover dummy classes leaking into namespace." + "reason": "Both are CI/repo hygiene issues, but they involve different files and unrelated breakages." }, { "left": "issue:44623", "right": "issue:45397", "accept": false, - "reason": "Processor save missing files and gemma-4 zero3 from_pretrained are different load/save defects." + "reason": "processor.save_pretrained missing files is unrelated to gemma-4 zero3 from_pretrained failures." }, { "left": "issue:44521", "right": "issue:45325", "accept": false, - "reason": "Multimodal assistant mask generation is unrelated to Qwen2.5-VL rope index scaling." + "reason": "Both are multimodal, but assistant_mask generation and rope index scaling are distinct bugs." }, { "left": "issue:33290", "right": "issue:45656", "accept": false, - "reason": "Both mention deepspeed/optimizer, but one is OOM with Adafactor and the other is optimizer.step being called twice." + "reason": "Both mention deepspeed and optimizers, but OOM in Adafactor and double optimizer stepping are different problems." }, { "left": "issue:44261", "right": "issue:44263", "accept": false, - "reason": "Missing rms_norm_eps precision issue is unrelated to torch.split return values in GlmMoeDsaIndexer." + "reason": "Different failures in different components: MLA layernorm precision vs torch.split return handling." }, { "left": "issue:46123", "right": "issue:46369", "accept": false, - "reason": "Final partial batch dropping results is not the same as memory overflow in sam3-litetext." + "reason": "MaskGenerationPipeline batch handling and sam3-litetext memory overflow are unrelated." }, { "left": "issue:44361", "right": "issue:45593", "accept": false, - "reason": "MLukeTokenizer AttributeError and D-FINE auxiliary-loss behavior are unrelated." + "reason": "MLukeTokenizer task errors do not match D-FINE auxiliary-loss behavior." }, { "left": "issue:44016", "right": "issue:45823", "accept": false, - "reason": "Notebook syntax error is not the same as Gemma4 PLE device mismatch during forward." + "reason": "A notebook syntax error and a Gemma4 device-mismatch forward bug are unrelated." }, { "left": "issue:43441", "right": "issue:44683", "accept": false, - "reason": "Both concern attention kernels, but one is Ministral FlashAttention failure and the other is compiled flex_attention on torch>=2.9." + "reason": "Both concern attention kernels, but they affect different kernels/models and have different failure modes." }, { "left": "issue:44671", "right": "issue:44779", "accept": false, - "reason": "CamemBERT masked-LM predictions and Deepseek tokenizer corruption are different model/tokenizer bugs." + "reason": "Both are v5 regressions, but CamemBERT MLM output and Deepseek tokenizer correctness are different subsystems." }, { "left": "issue:43901", "right": "issue:43994", "accept": false, - "reason": "Docs mentioning return_all_scores is unrelated to SigLIP2 producing bad outputs." + "reason": "Docs mismatch for return_all_scores is unrelated to incorrect model/pipeline outputs for SigLIP2." }, { "left": "issue:43723", "right": "issue:44162", "accept": false, - "reason": "AutoTokenizer loading regression and ESM2 breakage are separate model-loading issues." + "reason": "AutoTokenizer loading issues and ESM2 breakage are separate bugs." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "Both involve kernels, but one is a pipelines import dependency issue and the other is a Qwen3.5 registration crash." }, { "left": "issue:44297", "right": "issue:45397", "accept": false, - "reason": "Tokenizer save metadata mismatch is unrelated to gemma-4 zero3 from_pretrained failures." + "reason": "Tokenizer save metadata mismatch is unrelated to gemma-4 zero3 loading." }, { "left": "issue:43525", "right": "issue:43881", "accept": false, - "reason": "Llama4Config missing pad_token_id and glm-4v-9b loading failure are different model-specific bugs." + "reason": "Both are model-loading issues, but they involve different models and different missing-field/root causes." }, { "left": "issue:43901", "right": "issue:44246", "accept": false, - "reason": "Pipeline docs mismatch is unrelated to intermittent import-time slowness." + "reason": "Pipeline docs wording and slow import are unrelated." }, { "left": "issue:44016", "right": "issue:44291", "accept": false, - "reason": "A notebook syntax error does not match the init_empty_weights unexpected-argument TypeError." + "reason": "Notebook syntax error and init_empty_weights argument handling are unrelated." }, { "left": "issue:43525", "right": "issue:43643", "accept": false, - "reason": "Missing pad_token_id in Llama4Config and trust_remote_code missing fields are not the same defect." + "reason": "Both mention missing config fields, but one is a specific model config attribute and the other is a trust_remote_code config-return bug." }, { "left": "issue:44589", "right": "issue:45362", "accept": false, - "reason": "Float8 storage lookup failure is unrelated to Qwen3.5 chat crashes." + "reason": "Storage-object import failure and Qwen3.5 chat crash are unrelated." }, { "left": "issue:44188", "right": "issue:44336", "accept": false, - "reason": "torch.compile attention divergence and ANSI codes in loading_report are completely different bugs." + "reason": "Attention-kernel divergence and ANSI output formatting are unrelated." }, { "left": "issue:41628", "right": "issue:43976", "accept": false, - "reason": "AutoImageProcessor import failure is unrelated to the Python version compatibility issue." + "reason": "Import error for AutoImageProcessor and Python-version compatibility are different issues." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both are dependency/import problems, but sentence-transformers nn import and TensorBoard TF backend handling are different code paths." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Generation cache missing during chunked prefill is unrelated to CLIPVisionModel hidden_states regression." } ] }, @@ -11843,12 +11786,13 @@ "issue:43299", "issue:43352", "issue:43475", + "issue:43526", + "issue:43596", "issue:43673", - "issue:43761", - "issue:43784", "issue:43824", "issue:43901", "issue:43976", + "issue:44162", "issue:44220", "issue:44242", "issue:44246", @@ -11867,23 +11811,20 @@ "issue:44964", "issue:45003", "issue:45103", + "issue:45200", "issue:45250", "issue:45292", "issue:45362", "issue:45375", "issue:45397", - "issue:45542", "issue:45593", "issue:45859", "issue:46132", - "issue:46291", "issue:46369", "issue:46399", "issue:46424" ], "soft_pairs": [ - "issue:43784|issue:45542", - "issue:43673|issue:43761", "issue:44336|issue:44373", "issue:16998|issue:22355", "issue:45103|issue:46399", @@ -11904,222 +11845,219 @@ "issue:42886|issue:46132", "issue:44295|issue:44749", "issue:43475|issue:46369", - "issue:46291|issue:46399", "issue:45362|issue:45375", "issue:45859|issue:46399", "issue:44779|issue:45292", "issue:43673|issue:44561", "issue:43901|issue:45003", - "issue:43352|issue:45250" + "issue:43352|issue:45250", + "issue:44162|issue:45200", + "issue:36246|issue:43596", + "issue:43526|issue:45003", + "issue:44496|issue:45003" ], - "prepared_review_unit_hash": "39ba06f3726343ee3f6ca4f60d67e63de2a32dfb58bde10fa7c408d86724727b", + "prepared_review_unit_hash": "cce7cebeffe016de2a8dc96a3ea6fa15248be6b9b1c72202290b38150fdadb6b", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, - "soft_pair_count": 29, - "serialized_chars": 35689, - "estimated_input_tokens": 8923, - "estimated_eval_tokens": 18102 + "soft_pair_count": 30, + "serialized_chars": 35960, + "estimated_input_tokens": 8990, + "estimated_eval_tokens": 18236 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These are mostly separate Transformers issues across import regressions, model-loading failures, cache/attention bugs, and a few docs/runtime nits. None of the soft pairs look like the same underlying bug/change, so I would not merge any of them as duplicates.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 43824 is the strongest representative of the recurring v5 import/export regression theme: it is a concrete top-level missing-symbol failure, has multiple inbound references, and is a clear user-facing breakage.", + "summary": "This cluster is heterogeneous: the soft links mostly connect unrelated import errors, model-specific regressions, docs/perf issues, and cache/config bugs. I found no true duplicate pair to merge, and there is no PR content here.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items span many unrelated bugs and regressions, so none is a safe duplicate representative.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43784 is a very clean, reproducible downstream import failure with a precise error message; it is a good exemplar of the import-breakage class even though it is narrower than 43824.", + "best_issue_reason": "No single best issue candidate; this is not a coherent duplicate cluster, so choosing one issue would be misleading.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43784", - "right": "issue:45542", - "accept": false, - "reason": "Different failures: sentence-transformers import NameError vs TensorBoard/TensorFlow backend setup." - }, - { - "left": "issue:43673", - "right": "issue:43761", - "accept": false, - "reason": "Both are v5 regressions, but cache handling in chunked prefill and CLIP hidden_states behavior are different code paths." - }, { "left": "issue:44336", "right": "issue:44373", "accept": false, - "reason": "ANSI code output bug in loading_report vs a wrong docstring for position_ids; runtime vs docs." + "reason": "ANSI-code rendering in loading_report is unrelated to a docstring typo for position_ids." }, { "left": "issue:16998", "right": "issue:22355", "accept": false, - "reason": "DeBERTa model_max_length question is unrelated to the missing transformers.onnx module import error." + "reason": "model_max_length on DeBERTa-V3 is unrelated to the missing transformers.onnx module." }, { "left": "issue:45103", "right": "issue:46399", "accept": false, - "reason": "auto_docstring annotation crash and Qwen3.5 rotary-kernel registration crash are unrelated bugs." + "reason": "auto_docstring annotation parsing and Qwen3.5 rotary-kernel registration are different code paths and bugs." }, { "left": "issue:44242", "right": "issue:45593", "accept": false, - "reason": "MoE load-balancing loss omission vs D-FINE auxiliary-loss behavior; different models and loss logic." + "reason": "Missing load-balancing loss in MoE routing is not the same as D-FINE auxiliary-loss handling when denoising is off." }, { "left": "issue:44295", "right": "issue:45397", "accept": false, - "reason": "position_ids buffer access error is not the same as gemma-4 zero3 from_pretrained failure." + "reason": "A position_ids buffer access error is unrelated to Gemma-4 zero3 from_pretrained loading." }, { "left": "issue:44246", "right": "issue:44496", "accept": false, - "reason": "Import slowdown sometimes vs unrecognized model/config error; no shared underlying bug." + "reason": "Import latency for transformers is unrelated to an unrecognized model/config error." }, { "left": "issue:42491", "right": "issue:43299", "accept": false, - "reason": "Both mention Qwen3 MoE, but one is LoRA compatibility on hf4.x/hf5.x and the other is Qwen3VL model loading; different concrete failures." + "reason": "Both mention Qwen3 MoE, but one is LoRA compatibility across versions and the other is model loading breakage; not the same concrete bug." }, { "left": "issue:44246", "right": "issue:44871", "accept": false, - "reason": "Import timing issue vs Gemma eos_token_id configuration mismatch." + "reason": "Import performance and eos_token_id configuration mismatch are unrelated." }, { "left": "issue:44336", "right": "issue:44485", "accept": false, - "reason": "Loading-report ANSI escape codes vs GLM-5 RoPE implementation discussion are unrelated." + "reason": "Terminal ANSI output in loading_report has nothing to do with GLM-5 RoPE implementation." }, { "left": "issue:43824", "right": "issue:43976", "accept": false, - "reason": "Missing Qwen2_5_VL export is unrelated to Python 3.9/3.10 compatibility." + "reason": "A missing Qwen2_5_VL export is unrelated to Python-version compatibility for Transformers 5.1.0." }, { "left": "issue:41950", "right": "issue:44220", "accept": false, - "reason": "Video pipeline image-processor lookup and _torch_extract_fbank_features are separate audio/video code paths." + "reason": "Video-classification image-processor lookup is unrelated to _torch_extract_fbank_features()." }, { "left": "issue:44734", "right": "issue:46424", "accept": false, - "reason": "KV-cache continuation tensor indexing bug is different from max_cache_length being honored only for quantized cache." + "reason": "KV-cache continuation indexing in serve is different from max_cache_length semantics for Quantized Cache." }, { "left": "issue:36246", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but one is a missing Qwen2_5_VL symbol and the other is a huggingface_hub API import break." + "reason": "A missing Qwen2_5_VL image processor export is unrelated to huggingface_hub's is_offline_mode import." }, { "left": "issue:36246", "right": "issue:44351", "accept": false, - "reason": "Missing Qwen2_5_VLImageProcessor export vs missing HybridCache export; same symptom class, different symbols/causes." - }, - { - "left": "issue:44336", - "right": "issue:46424", - "accept": false, - "reason": "ANSI-code rendering bug is unrelated to cache-length configuration behavior." + "reason": "Different missing symbols from different modules; not the same import bug." }, { "left": "issue:43976", "right": "issue:44964", "accept": false, - "reason": "Python-version support regression vs a specific Phi-4 multimodal loading failure." + "reason": "General Python-version support failure is unrelated to a specific Phi-4 multimodal loading error." }, { "left": "issue:44242", "right": "issue:44857", "accept": false, - "reason": "MoE load-balancing loss issue is unrelated to the LwDetrImageLoss AMP/CUDA crash." + "reason": "MoE load-balancing loss logic and LWDETR AMP crash are different failures." }, { "left": "issue:42886", "right": "issue:46132", "accept": false, - "reason": "Offline cache loading failure and AttentionInterface.register behavior change do not describe the same bug." + "reason": "Offline cache loading and AttentionInterface.register behavior are unrelated." }, { "left": "issue:44295", "right": "issue:44749", "accept": false, - "reason": "Buffer access error and tokenizer slowdown are unrelated regressions." + "reason": "A buffer-related position_ids error is not the same as slow filtering after upgrading Transformers." }, { "left": "issue:43475", "right": "issue:46369", "accept": false, - "reason": "SAM3 vision-encoder missing attribute and SAM3 litetext memory overflow are different failure modes." - }, - { - "left": "issue:46291", - "right": "issue:46399", - "accept": false, - "reason": "Kernels v0.15 version/revision requirement for evaluate is unrelated to Qwen3.5 apply_rotary_pos_emb registration failure." + "reason": "SAM3 encoder output missing an attribute is unrelated to memory overflow in sam3-litetext." }, { "left": "issue:45362", "right": "issue:45375", "accept": false, - "reason": "Qwen3.5 chat crash and missing deepstack_visual_indexes in config are different issues." + "reason": "A chat crash and a missing config field may be related at a high level, but they are not clearly the same concrete bug." }, { "left": "issue:45859", "right": "issue:46399", "accept": false, - "reason": "CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding is unrelated to the missing decorator crash in kernels registration." + "reason": "CPU-offload incompatibility in text rotary embedding is different from the missing decorator error during apply_rotary_pos_emb registration." }, { "left": "issue:44779", "right": "issue:45292", "accept": false, - "reason": "DeepSeek tokenizer regression and resize_token_embeddings not updating output embeddings are different subsystems." + "reason": "Deepseek tokenizer correctness and resize_token_embeddings output embedding syncing are different issues." }, { "left": "issue:43673", "right": "issue:44561", "accept": false, - "reason": "Cache missing during chunked prefill is unrelated to removal of is_torch_fx_available breaking trust_remote_code models." + "reason": "Chunked-prefill cache absence is unrelated to removal of is_torch_fx_available breaking trust_remote_code models." }, { "left": "issue:43901", "right": "issue:45003", "accept": false, - "reason": "Docs wording for return_all_scores does not match unsafe sys.modules access in modeling_utils." + "reason": "A stale docstring about return_all_scores is unrelated to unsafe sys.modules access in modeling_utils." }, { "left": "issue:43352", "right": "issue:45250", "accept": false, - "reason": "Nemotron lacking Flash Attention 2 support is not the same as the generic Flash Attention 2 issue." + "reason": "Flash-Attention support for Nemotron and a generic Flash Attention topic are not the same bug." }, { - "left": "issue:44749", - "right": "issue:45292", + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma-4 mm_token_type_ids defaults are unrelated model-specific bugs." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "A Qwen2_5_VL import error is unrelated to a DeepSpeed zero3/BertModel IndexError." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and modeling_utils sys.modules access are different code paths." + }, + { + "left": "issue:44496", + "right": "issue:45003", "accept": false, - "reason": "Tokenizer slowdown after upgrade and resize_token_embeddings/output_embeddings mismatch are unrelated." + "reason": "Unrecognized model/config loading and unsafe sys.modules access are unrelated problems." } ] }, @@ -12143,24 +12081,23 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:36246", "issue:43065", "issue:43257", "issue:43262", "issue:43408", "issue:43475", - "issue:43526", "issue:43540", - "issue:43596", "issue:43653", "issue:43906", + "issue:43931", + "issue:44060", + "issue:44077", "issue:44162", "issue:44246", "issue:44368", "issue:44466", "issue:44479", "issue:44483", - "issue:44496", "issue:44561", "issue:44877", "issue:44977", @@ -12193,10 +12130,6 @@ "issue:46431" ], "soft_pairs": [ - "issue:44162|issue:45200", - "issue:36246|issue:43596", - "issue:43526|issue:45003", - "issue:44496|issue:45003", "issue:43257|issue:45137", "issue:45292|issue:45584", "issue:45412|issue:46123", @@ -12221,210 +12154,214 @@ "issue:44368|issue:45563", "issue:43065|issue:43408", "issue:43906|issue:46202", - "issue:45987|issue:46082" + "issue:45987|issue:46082", + "issue:43931|issue:44368", + "issue:44877|issue:46170", + "issue:44060|issue:44077", + "issue:45003|issue:45561" ], - "prepared_review_unit_hash": "bf8f33755dacb50a2e2f82554c2500d43264322ae89e8c30accb20262a3a84a6", + "prepared_review_unit_hash": "de55b7499700e20b9563dad6bd201fb3319dc385df64dcebcb0073a25b64f761", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, + "node_count": 47, + "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 36627, - "estimated_input_tokens": 9157, - "estimated_eval_tokens": 18570 + "serialized_chars": 35873, + "estimated_input_tokens": 8969, + "estimated_eval_tokens": 18194 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Loose collection of unrelated bug reports across model loading, configs, tokenizers, video/audio processors, and testing infra. None of the soft-similarity pairs look like the same underlying bug/change.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44479 is the broadest regression-style report in the set and the best umbrella candidate, but the cluster does not actually contain a true duplicate set.", + "summary": "The cluster is highly heterogeneous: all soft links are broad topical similarities, not clear duplicates of the same bug or code change. I would reject every soft edge and avoid choosing a single canonical artifact for the set.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items cover unrelated failures across SAM3, Qwen, Gemma, DeepSpeed, tokenization, config loading, and test infrastructure.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44479 has the widest apparent blast radius across multiple Qwen multimodal variants, making it the most representative issue despite weak overall cluster cohesion.", + "best_issue_reason": "No single issue is a strong global representative because there is no coherent duplicate cluster; the titles point to distinct root causes and subsystems.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44162", - "right": "issue:45200", - "accept": false, - "reason": "Different domains: ESM2 model breakage vs Gemma 4 multimodal token defaults." - }, - { - "left": "issue:36246", - "right": "issue:43596", - "accept": false, - "reason": "Unrelated failures: Qwen2.5-VL import error vs DeepSpeed/Bert zero3 index error." - }, - { - "left": "issue:43526", - "right": "issue:45003", - "accept": false, - "reason": "Different bugs in different code paths: BEiT label reduction vs unsafe sys.modules access." - }, - { - "left": "issue:44496", - "right": "issue:45003", - "accept": false, - "reason": "Model config loading error vs internal sys.modules lookup bug; not the same issue." - }, { "left": "issue:43257", "right": "issue:45137", "accept": false, - "reason": "Qwen3 MOE weight conversion under accelerate/deepspeed is unrelated to ZeRO3 deque underflow." + "reason": "Both mention DeepSpeed, but one is a Qwen3 MOE conversion bug and the other is a ZeRO3 deque crash; different code paths and symptoms." }, { "left": "issue:45292", "right": "issue:45584", "accept": false, - "reason": "Different subsystems: token embedding resize behavior vs Whisper empty transcription generation." + "reason": "Resize-token-embeddings propagation and Whisper empty-transcription handling are unrelated bugs." }, { "left": "issue:45412", "right": "issue:46123", "accept": false, - "reason": "RT-DETR memory release bug is unrelated to MaskGenerationPipeline batch handling." + "reason": "Memory release after RT-DETR deletion is unrelated to MaskGenerationPipeline batch-finalization behavior." }, { "left": "issue:43262", "right": "issue:44246", "accept": false, - "reason": "Audio chat-template sampling rate default bug vs import-time slowdown are different problems." + "reason": "Audio sampling-rate defaults and slow import are different issues with no shared root cause." }, { "left": "issue:44561", "right": "issue:46399", "accept": false, - "reason": "Removal of is_torch_fx_available breaks trust_remote_code models, while the other is a rotary-pos-emb registration issue." + "reason": "One is a v5 compatibility break in modeling_utils; the other is a Qwen3.5 kernel registration crash." }, { "left": "issue:45292", "right": "issue:45701", "accept": false, - "reason": "resize_token_embeddings behavior does not match a broad tokenization-version-change report." + "reason": "Embedding resize propagation and tokenization changes across versions are not the same bug." }, { "left": "issue:45375", "right": "issue:45759", "accept": false, - "reason": "Missing config field in Qwen3.5MoeVisionConfig vs composite config unwrapping in AutoModelForCausalLM." + "reason": "Both involve Qwen multimodal config handling, but one is a missing config field under strict mode and the other is incorrect text_config unwrapping; not the same failure." }, { "left": "issue:44877", "right": "issue:45375", "accept": false, - "reason": "Strict config loading for granite_speech is unrelated to Qwen3.5MoeVisionConfig field dropping." + "reason": "Strict config loading breaks different model families for different reasons; these are not duplicates." }, { "left": "issue:45561", "right": "issue:45563", "accept": false, - "reason": "Testing-utils xdist race condition is unrelated to paged generate stale warning behavior." + "reason": "A pytest-xdist file race and a stale generate() warning are unrelated." }, { "left": "issue:44987", "right": "issue:45459", "accept": false, - "reason": "Loading physical-intelligence/fast on newer transformers is not the same as tokenizer decode-error masking." + "reason": "Model-load failure on a specific repo and protobuf-related tokenizer error masking are different underlying bugs." }, { "left": "issue:44466", "right": "issue:45335", "accept": false, - "reason": "Both involve embeddings, but one is tied-weight serialization and the other is t5gemma resize propagation." + "reason": "Both concern tied/embedding weights, but one is serialization inconsistency and the other is resize_token_embeddings not updating a decoder embedding." }, { "left": "issue:44995", "right": "issue:45200", "accept": false, - "reason": "GlmMoeDsa second-forward stale cache bug is unrelated to Gemma 4 mm_token_type_ids defaults." + "reason": "A stale cache crash in GlmMoeDsa is unrelated to missing mm_token_type_ids defaults." }, { "left": "issue:45397", "right": "issue:45656", "accept": false, - "reason": "Gemma 4 ZeRO3 loading issue is not the same as optimizer stepping twice under deepspeed." + "reason": "Gemma-4 ZeRO3 loading and duplicate optimizer stepping under deepspeed are different problems." }, { "left": "issue:45137", "right": "issue:45600", "accept": false, - "reason": "Different DeepSpeed-related failures: deque underflow vs stale auto_mappings references." + "reason": "A ZeRO3 deque error and stale auto_mappings references are separate CI/runtime issues." }, { "left": "issue:45685", "right": "issue:46431", "accept": false, - "reason": "MPS histogram-on-Int MoE error is unrelated to missing chat-template support." + "reason": "An MPS histogram integer-support error and missing chat template support are unrelated." }, { "left": "issue:43475", "right": "issue:43540", "accept": false, - "reason": "SAM3 vision encoder attribute error vs Qwen3OmniMoe video ValueError are different model bugs." + "reason": "Different SAM3/Qwen video-processing bugs with different object/model failures." }, { "left": "issue:44479", "right": "issue:46170", "accept": false, - "reason": "Qwen multimodal video regression is unrelated to Windows/ROCm distributed import/type errors." + "reason": "A v5.3 video regression and a Windows ROCm import/dependency issue are unrelated." }, { "left": "issue:43653", "right": "issue:44162", "accept": false, - "reason": "Tokenizer special-token registration bug is unrelated to ESM2 model breakage." + "reason": "Tokenizer special-token registration and ESM2 breakage are not the same defect." }, { "left": "issue:44483", "right": "issue:46202", "accept": false, - "reason": "Chat completions request validation regression is unrelated to Gemma pipeline roundtrip tokenization." + "reason": "An API request rejection and a Gemma roundtrip tokenization failure are different bugs." }, { "left": "issue:44368", "right": "issue:44977", "accept": false, - "reason": "A tie_word_embeddings warning from fine-tuning is not the same as a flash-attention generation failure." + "reason": "Both are Qwen3.5-related, but one is a warning about tied embeddings and the other is a flash-attention generation failure." }, { "left": "issue:44995", "right": "issue:45588", "accept": false, - "reason": "Stale cache on second forward pass is unrelated to flash_attention.py crashing on s_aux=None." + "reason": "GlmMoeDsa stale cache crash and flash_attention s_aux=None AttributeError are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A config warning and a paged-generation stale warning are not the same code-path failure." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 pixel-decoder dummy conv and model-type mismatch warning are different SAM3 issues." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "Isolated reproduction of an older issue and a Gemma pipeline roundtrip bug are unrelated." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "StaticCache return-type contract and LlamaConfig head_dim validation are distinct model-internal issues." }, { - "left": "issue:44368", - "right": "issue:45563", + "left": "issue:43931", + "right": "issue:44368", "accept": false, - "reason": "Model config warning during LoRA fine-tuning is unrelated to paged generate stale warnings." + "reason": "Qwen3-VL weight-shape mismatch and Qwen3.5 tied-weights warning do not share a concrete root cause." }, { - "left": "issue:43065", - "right": "issue:43408", + "left": "issue:44877", + "right": "issue:46170", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder and sampler type warning involve different SAM3 components." + "reason": "Strict config loading for granite_speech and Windows ROCm/torch.distributed import issues are unrelated." }, { - "left": "issue:43906", - "right": "issue:46202", + "left": "issue:44060", + "right": "issue:44077", "accept": false, - "reason": "Repro of an earlier issue is unrelated to Gemma pipeline roundtrip failure." + "reason": "Qwen3-Next tied-weights warning and patchtsmixer post_init policy are not the same bug." }, { - "left": "issue:45987", - "right": "issue:46082", + "left": "issue:45003", + "right": "issue:45561", "accept": false, - "reason": "StaticCache seq-length return-type contract issue is unrelated to LlamaConfig head_dim validation." + "reason": "Unsafe sys.modules access and an xdist file race are separate infrastructure problems." } ] }, @@ -12452,6 +12389,7 @@ "issue:30990", "issue:41628", "issue:42371", + "issue:42673", "issue:42886", "issue:42994", "issue:43232", @@ -12462,24 +12400,24 @@ "issue:43646", "issue:43673", "issue:43701", + "issue:43704", "issue:43784", - "issue:43931", "issue:44038", - "issue:44060", - "issue:44077", "issue:44162", "issue:44188", "issue:44297", + "issue:44351", "issue:44361", - "issue:44368", "issue:44623", "issue:44749", "issue:44869", "issue:44871", - "issue:44877", "issue:44933", - "issue:45003", + "issue:44964", + "issue:45081", "issue:45092", + "issue:45103", + "issue:45137", "issue:45229", "issue:45230", "issue:45310", @@ -12488,7 +12426,6 @@ "issue:45381", "issue:45397", "issue:45478", - "issue:45561", "issue:45584", "issue:45656", "issue:46132", @@ -12498,10 +12435,6 @@ "issue:46421" ], "soft_pairs": [ - "issue:43931|issue:44368", - "issue:44877|issue:46170", - "issue:44060|issue:44077", - "issue:45003|issue:45561", "issue:42886|issue:45229", "issue:45584|issue:46170", "issue:44297|issue:44623", @@ -12526,216 +12459,220 @@ "issue:45375|issue:45478", "issue:42994|issue:43525", "issue:45310|issue:45375", - "issue:43784|issue:46399" + "issue:43784|issue:46399", + "issue:42673|issue:43704", + "issue:45081|issue:45103", + "issue:44964|issue:45137", + "issue:41628|issue:44351" ], - "prepared_review_unit_hash": "b8e4b94b2761ca5f135eb24bdf9e79f0635319d47ace9db5a06ac05bd72cca15", + "prepared_review_unit_hash": "c67c11a63efebc0efd1fceba07a5ff4de7a99379b66ce030fc035ded4c8a6b0d", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 29, - "serialized_chars": 36170, - "estimated_input_tokens": 9043, - "estimated_eval_tokens": 18342 + "serialized_chars": 36204, + "estimated_input_tokens": 9051, + "estimated_eval_tokens": 18358 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a heterogeneous bucket of unrelated Transformers bugs. The only clear duplicate chain is the Qwen3.5-Moe `from_pretrained` regression: issue 45310 is the broad report, 45375 narrows it to the missing `deepstack_visual_indexes` field, and 45478 is a re-report of the same failure.", - "confidence": 0.86, - "canonical_issue_reason": "Issue 45310 is the broadest, most general report of the Qwen3.5-Moe `from_pretrained` regression; 45375 provides the concrete root cause, and 45478 appears to be the same bug reported again.", + "summary": "Mostly a heterogeneous cluster of unrelated Transformers issues, with only two clear duplicate subclusters standing out: the Qwen3.5 Moe from_pretrained regression and the repeated Qwen3ForCausalLM VRAM leak.", + "confidence": 0.58, + "canonical_issue_reason": "Issue 45375 is the best anchor here because it gives the concrete root cause for the Qwen3.5 Moe from_pretrained regression and best subsumes the generic 45310/45478 reports.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 45310 is the best canonical issue because it captures the user-visible failure without being overly narrow, making it the most suitable anchor for the duplicate set.", + "best_issue_reason": "45375 is the strongest issue to keep as canonical because it explains the regression instead of only restating the symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43931", - "right": "issue:44368", - "accept": false, - "reason": "Different failures: a Qwen3-VL weight-shape load error vs a tie-weights warning for Qwen3.5 fine-tuning." - }, - { - "left": "issue:44877", - "right": "issue:46170", - "accept": false, - "reason": "Unrelated: strict config loading for `granite_speech` vs Windows/ROCm import and `torch.distributed` dependency issues." - }, - { - "left": "issue:44060", - "right": "issue:44077", - "accept": false, - "reason": "Both touch model internals, but one is a tied-weights warning and the other is a `post_init` API policy change; not the same bug." - }, - { - "left": "issue:45003", - "right": "issue:45561", - "accept": false, - "reason": "Different code paths: unsafe `sys.modules` access vs a pytest-xdist race on `captured_info.txt`." - }, { "left": "issue:42886", "right": "issue:45229", "accept": false, - "reason": "Completely different problems: offline tokenizer cache loading vs multi-GPU CUDA OOM." + "reason": "Different bugs: offline cache tokenizer loading vs multi-GPU inference OOM." }, { "left": "issue:45584", "right": "issue:46170", "accept": false, - "reason": "Unrelated: Whisper empty-transcription generation failure vs Windows/ROCm import-time dependency issues." + "reason": "Unrelated failures: Whisper decoding crash vs Windows ROCm torch.distributed import error." }, { "left": "issue:44297", "right": "issue:44623", "accept": false, - "reason": "Both involve save/pretrained serialization, but tokenizer class mismatch and missing processor files are different concrete bugs." + "reason": "Both involve save_pretrained, but one is tokenizer metadata and the other is processor file output." }, { "left": "issue:44361", "right": "issue:44749", "accept": false, - "reason": "Different symptoms and subsystems: MLukeTokenizer AttributeError vs preprocessing slowdown after an upgrade." + "reason": "Different symptoms: tokenizer AttributeError vs a broad speed regression." }, { "left": "issue:44749", "right": "issue:45656", "accept": false, - "reason": "Filtering slowdown and optimizer stepping twice under deepspeed are unrelated behaviors." + "reason": "Performance regression vs optimizer-step behavior; not the same code path." }, { "left": "issue:44623", "right": "issue:45656", "accept": false, - "reason": "Processor save/pretrained file issues do not match deepspeed optimizer double-step behavior." + "reason": "Processor save_pretrained bug vs optimizer double-step bug." }, { "left": "issue:43701", "right": "issue:44749", "accept": false, - "reason": "Checkpoint resume key mismatch is unrelated to preprocessing latency." + "reason": "Checkpoint resume key mismatch vs model loading slowdown." }, { "left": "issue:43646", "right": "issue:44162", "accept": false, - "reason": "`trust_remote_code` missing fields and ESM2 breakage are different failures and likely different fixes." + "reason": "AutoConfig field-loss bug vs ESM2 model regression." }, { "left": "issue:44869", "right": "issue:45356", "accept": false, - "reason": "Whisper timestamp decode crash and Kimi tokenizer codec regression are distinct tokenizer/decoding bugs." + "reason": "Whisper timestamp decode crash vs Kimi tokenizer regression." }, { "left": "issue:44871", "right": "issue:46132", "accept": false, - "reason": "Gemma eos-token-id configuration mismatch is unrelated to AttentionInterface registration behavior changes." + "reason": "Gemma eos_token_id mismatch vs AttentionInterface behavior change." }, { "left": "issue:45230", "right": "issue:45397", "accept": false, - "reason": "Generic bug report vs a specific Gemma-4 zero3 `from_pretrained` issue; not enough to be the same bug." + "reason": "Generic bug report vs specific Gemma-4 zero3 from_pretrained failure." }, { "left": "issue:44188", "right": "issue:46202", "accept": false, - "reason": "Attention-kernel divergence under `torch.compile` is unrelated to Gemma pipeline roundtrip behavior." + "reason": "Attention kernel divergence under torch.compile vs Gemma roundtrip token issue." }, { "left": "issue:42994", "right": "issue:43482", "accept": false, - "reason": "Quantized model saving failure and Qwen2.5-GGUF loading failure are different quantization/load paths." + "reason": "Quantized model saving failure vs GGUF loading failure." }, { "left": "issue:44038", "right": "issue:44623", "accept": false, - "reason": "Qwen3-VL-Moe / transformers 5 behavior is unrelated to processor save/pretrained missing files." + "reason": "Qwen3-VL-Moe model issue vs processor save_pretrained files." }, { "left": "issue:43475", "right": "issue:43643", "accept": false, - "reason": "SAM3 vision encoder attribute error and `trust_remote_code` config field loss are different bugs." + "reason": "SAM3 encoder output attribute error vs AutoConfig field-loss bug." }, { "left": "issue:28282", "right": "issue:30990", "accept": false, - "reason": "AutoModel missing PyTorch import and Sentence Transformers loading hang are not the same underlying issue." + "reason": "AutoModel missing PyTorch vs sentence-transformers loading stall." }, { "left": "issue:42371", "right": "issue:45092", "accept": false, - "reason": "TF32 API deprecation is unrelated to InternVL2 meta-initialization incompatibility." + "reason": "TF32 API migration issue vs InternVL2 meta-initialization incompatibility." }, { "left": "issue:43475", "right": "issue:45381", "accept": false, - "reason": "SAM3 vision encoder missing attribute vs Qwen2.5-VL video position IDs bug are unrelated." + "reason": "SAM3 encoder output error vs qwen2.5-vl video position ids bug." }, { "left": "issue:41628", "right": "issue:44933", "accept": false, - "reason": "Both mention import/export problems, but they refer to different missing names and likely different breakages." + "reason": "Missing AutoImageProcessor import vs unrelated image_utils import issue." }, { "left": "issue:43232", "right": "issue:43673", "accept": false, - "reason": "Generation kwargs sync after `sync_gpus` and missing cache during chunked prefill are different generation-path bugs." + "reason": "Generation kwarg update bug vs cache missing during chunked prefill; distinct paths." }, { "left": "issue:44188", "right": "issue:46421", "accept": false, - "reason": "Different cache/compile issues: divergent attention kernels vs spurious recompilation with chunked prefill." + "reason": "torch.compile attention branching issue vs cache lazy-init recompilation issue." }, { "left": "issue:45375", "right": "issue:45478", "accept": true, - "reason": "Same Qwen3.5-Moe `from_pretrained` failure on transformers>=5.4.0; 45478 is a re-report of the same bug." + "reason": "Same Qwen3.5 Moe from_pretrained regression; 45375 is the root-cause report." }, { "left": "issue:42994", "right": "issue:43525", "accept": false, - "reason": "Quantized save failure and Llama4 missing `pad_token_id` are unrelated." + "reason": "Quantized save bug vs Llama4 config missing pad_token_id." }, { "left": "issue:45310", "right": "issue:45375", "accept": true, - "reason": "Same Qwen3.5-Moe load regression; 45375 is the concrete missing-field diagnosis of the failure reported in 45310." + "reason": "Same bug family: generic Qwen3.5 Moe from_pretrained failure and the missing-field root cause." }, { "left": "issue:43784", "right": "issue:46399", "accept": false, - "reason": "Sentence-transformers import `nn` failure and Qwen3.5 kernel registration error are unrelated." + "reason": "sentence-transformers import nn error vs Qwen3.5 kernel registration error." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact duplicate Qwen3ForCausalLM VRAM leak report." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Mistral regex patch crash vs auto_docstring future-annotations crash." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal load failure vs DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Missing AutoImageProcessor import vs HybridCache import error." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is well grounded in the packet and stays conservative on the soft-edge pairs. The Qwen3.5-Moe chain (45310 \u2192 45375 \u2192 45478) is supported by matching titles and the later diagnosis issue, while the other pair verdicts avoid overclaiming same-bug relationships. No obvious overstatement found." + "feedback": "The analysis is well grounded in the packet and stays conservative on the soft edges. The accepted pairs are plausible duplicates/same-bug subclusters (45310/45375, 45375/45478, 42673/43704), and the rejects are generally well justified by distinct symptoms or code paths. The summary is appropriately cautious in describing the cluster as mostly heterogeneous with only those two clear duplicate subclusters." }, "evaluator_used": true, "retried": false, @@ -12764,7 +12701,6 @@ "issue:41628", "issue:42175", "issue:42222", - "issue:42673", "issue:42831", "issue:42898", "issue:42994", @@ -12772,28 +12708,29 @@ "issue:43232", "issue:43278", "issue:43377", - "issue:43704", "issue:43723", "issue:43784", "issue:43824", + "issue:43874", "issue:43901", "issue:44016", "issue:44038", "issue:44188", "issue:44222", "issue:44242", - "issue:44351", "issue:44373", "issue:44496", "issue:44568", "issue:44623", + "issue:44734", + "issue:44779", "issue:44792", - "issue:44964", - "issue:45081", + "issue:45092", "issue:45103", "issue:45137", "issue:45200", "issue:45310", + "issue:45356", "issue:45406", "issue:45561", "issue:45600", @@ -12805,10 +12742,6 @@ "issue:46231" ], "soft_pairs": [ - "issue:42673|issue:43704", - "issue:45081|issue:45103", - "issue:44964|issue:45137", - "issue:41628|issue:44351", "issue:45406|issue:46202", "issue:44222|issue:45310", "issue:42994|issue:43824", @@ -12833,220 +12766,221 @@ "issue:42898|issue:45812", "issue:43723|issue:44568", "issue:44016|issue:46231", - "issue:43784|issue:43824" + "issue:43784|issue:43824", + "issue:43874|issue:45092", + "issue:44734|issue:45103", + "issue:42898|issue:44779", + "issue:42898|issue:45356" ], - "prepared_review_unit_hash": "87523177b787b4355983a6fd85d4c689b8596a2293d474d80ae945253eb2f0f0", + "prepared_review_unit_hash": "2b131d7b373c37a31ed6e4ca9558186c7231d6bc6bdc72a670229f96693e0ded", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 47, "item_count": 47, "soft_pair_count": 29, - "serialized_chars": 35801, - "estimated_input_tokens": 8951, - "estimated_eval_tokens": 18158 + "serialized_chars": 35867, + "estimated_input_tokens": 8967, + "estimated_eval_tokens": 18190 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly heterogeneous issue reports. The only clear duplicate pair is the repeated Qwen3ForCausalLM VRAM leak report; all other soft pairs appear to describe different bugs, models, or subsystems.", - "confidence": 0.96, - "canonical_issue_reason": "issue:42673 is the best canonical issue because it is the earlier, more substantial instance of the exact duplicate bug report also seen in issue:43704 (same title and same Qwen3ForCausalLM multi-dataloader-thread VRAM leak).", + "summary": "This cluster is mostly a set of superficially similar but substantively different bug reports; I would not merge any of the soft-edge pairs. The closest thematic overlap is among tokenizer/v5 regressions, but they still look like distinct code paths and fixes.", + "confidence": 0.79, + "canonical_issue_reason": "Issue #42898 is the clearest standalone regression report in the set and the most plausible hub for the tokenizer/v5 subset, but the overall cluster is too heterogeneous to treat as one duplicate group.", "canonical_pr_reason": null, - "best_issue_reason": "issue:42673 is the strongest representative of the cluster\u2019s only true duplicate pair: it is the same concrete bug as issue:43704 and is more likely to contain the fuller original context.", + "best_issue_reason": "Issue #42898 is the strongest representative issue: it is concrete, reproducible, and narrowly scoped to a specific tokenizer behavior regression.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42673", - "right": "issue:43704", - "accept": true, - "reason": "Exact same bug report and same title; both describe Qwen3ForCausalLM leaking VRAM when used in multiple dataloader threads." - }, - { - "left": "issue:45081", - "right": "issue:45103", - "accept": false, - "reason": "Different crash sites and causes: Mistral tokenizer regex patching vs auto-docstring processing with future annotations." - }, - { - "left": "issue:44964", - "right": "issue:45137", - "accept": false, - "reason": "Unrelated failures: Phi-4 multimodal model loading vs DeepSpeed ZeRO3 deque pop error." - }, - { - "left": "issue:41628", - "right": "issue:44351", - "accept": false, - "reason": "Both are import errors, but for different symbols and likely different fixes: AutoImageProcessor vs HybridCache." - }, { "left": "issue:45406", "right": "issue:46202", "accept": false, - "reason": "Different Gemma-related bugs: transformers serve processor attribute error vs pipeline roundtrip tokenization issue." + "reason": "Different bugs: Gemma4Processor missing `_tokenizer` vs Gemma pipeline roundtrip on `\u2581`." }, { "left": "issue:44222", "right": "issue:45310", "accept": false, - "reason": "Distinct FP8/Qwen issues: save_pretrained for moe models vs from_pretrained error for Qwen3.5 MoE." + "reason": "Different code paths: FP8 `save_pretrained` in MoE vs `from_pretrained` failure for Qwen3.5 MoE." }, { "left": "issue:42994", "right": "issue:43824", "accept": false, - "reason": "Different problems: quantized model saving failure vs missing Qwen2_5_VLForConditionalGeneration import." + "reason": "Unrelated failures: quantized model saving vs an import error for `Qwen2_5_VLForConditionalGeneration`." }, { "left": "issue:46077", "right": "issue:46123", "accept": false, - "reason": "Different code paths: Gemma4 attention mask construction vs MaskGenerationPipeline batching logic." + "reason": "Different functionality: attention-mask construction vs batch-finalization in `MaskGenerationPipeline`." }, { "left": "issue:44038", "right": "issue:45310", "accept": false, - "reason": "Related model family, but not the same concrete bug: Qwen3-VL-Moe on transformers 5.0 vs Qwen3.5 MoE from_pretrained on >=5.4.0." + "reason": "Both mention Qwen MoE, but the concrete failures differ and do not look like one fixable code-path bug." }, { "left": "issue:36010", "right": "issue:41628", "accept": false, - "reason": "Separate missing-import issues for different APIs and modules." + "reason": "Both are import errors, but for different symbols (`GenerationMixin` vs `AutoImageProcessor`) and likely different causes." }, { "left": "issue:42831", "right": "issue:44222", "accept": false, - "reason": "FP8 accuracy regression is not the same bug as FP8 save_pretrained failure." + "reason": "Accuracy regression in FineGrainedFP8 is not the same as a `save_pretrained` serialization failure." }, { "left": "issue:42175", "right": "issue:45712", "accept": false, - "reason": "Different packaging/CI issues: missing tensorflow backend in torch install vs leftover dummy classes leaking into repo checks." + "reason": "Packaging/backend dependency issue vs dummy-class cleanup/check-repo failure; not the same bug." }, { "left": "issue:44496", "right": "issue:45561", "accept": false, - "reason": "Unrecognized model/config loading error vs pytest-xdist file race; unrelated." + "reason": "Model config recognition failure and an xdist file race are unrelated." }, { "left": "issue:45561", "right": "issue:45600", "accept": false, - "reason": "Testing utility race condition vs stale auto_mappings references; different failures." + "reason": "Test-worker race on `captured_info.txt` vs stale auto-mapping references; different root causes." }, { "left": "issue:43901", "right": "issue:45561", "accept": false, - "reason": "Documentation mismatch in TextClassificationPipeline vs concurrent file access race in tests." + "reason": "Docs mismatch for pipeline return values vs a parallel test-file race." }, { "left": "issue:43232", "right": "issue:44792", "accept": false, - "reason": "Generation kwargs bug after sync_gpus is unrelated to the janus image generation test failure." + "reason": "Generation kwargs update after `sync_gpus` is unrelated to the Janus image-generation test failure." }, { "left": "issue:43377", "right": "issue:44242", "accept": false, - "reason": "Different MoE-related issues: padding-mask inconsistency in MIMI encoder vs missing load-balancing loss when router logits are disabled." + "reason": "Different model bugs: missing padding-mask support vs missing load-balancing loss when router logits are disabled." }, { "left": "issue:30990", "right": "issue:35141", "accept": false, - "reason": "Sentence Transformers loading hang and embedding reinitialization are different bugs." + "reason": "A loading hang and an embedding reinitialization bug are not the same underlying issue." }, { "left": "issue:43278", "right": "issue:45137", "accept": false, - "reason": "Embedding dtype drift during eval is unrelated to a ZeRO3 deque error." + "reason": "Embedding dtype drift in eval is unrelated to the DeepSpeed ZeRO3 deque underflow." }, { "left": "issue:42898", "right": "issue:44568", "accept": false, - "reason": "Different tokenizer/v5 behavior changes: clean_up_tokenization_spaces vs add_special_tokens not adding BOS/EOS." + "reason": "Both are tokenizer regressions, but one is cleanup behavior in v5 and the other is BOS/EOS insertion for a specific tokenizer." }, { "left": "issue:44623", "right": "issue:45310", "accept": false, - "reason": "Processor save_pretrained missing files is unrelated to Qwen3.5 MoE loading errors." + "reason": "`processor.save_pretrained(...)` missing files is a save/export issue, not the same as a Qwen MoE load failure." }, { "left": "issue:30333", "right": "issue:33290", "accept": false, - "reason": "MLflow reporting on failed training jobs is unrelated to a DeepSpeed AdaFactor OOM." + "reason": "MLflow reporting on failed training jobs is unrelated to an AdaFactor OOM in DeepSpeed." }, { "left": "issue:44188", "right": "issue:44373", "accept": false, - "reason": "Attention-kernel divergence under torch.compile is unrelated to a docstring typo for position_ids." + "reason": "Attention-kernel divergence and a wrong docstring are completely different kinds of changes." }, { "left": "issue:43065", "right": "issue:46123", "accept": false, - "reason": "Dummy Conv2d in Sam3PixelDecoder is unrelated to partial-batch result dropping in MaskGenerationPipeline." + "reason": "Dummy `Conv2d` cleanup in Sam3PixelDecoder is unrelated to the partial-batch `is_last` bug." }, { "left": "issue:42222", "right": "issue:45200", "accept": false, - "reason": "Different Gemma issues: broken vitpose models vs mm_token_type_ids defaults for text-only fine-tuning." + "reason": "Broken VitPose models and Gemma4 `mm_token_type_ids` defaults are different model-specific problems." }, { "left": "issue:42898", "right": "issue:45812", "accept": false, - "reason": "Different tokenizer regressions: cleanup spacing behavior vs Granite token IDs being wrong." + "reason": "Both are tokenizer regressions, but they affect different models and token-ID logic, so they are not one duplicate." }, { "left": "issue:43723", "right": "issue:44568", "accept": false, - "reason": "Tokenizer loading issue in v5 is not the same as add_special_tokens BOS/EOS behavior regression." + "reason": "Tokenizer loading in v5 and missing BOS/EOS on `add_special_tokens=True` are different behaviors." }, { "left": "issue:44016", "right": "issue:46231", "accept": false, - "reason": "Notebook syntax error and doc parameter-name mismatch are different documentation issues." + "reason": "Notebook syntax error and a docs parameter-name mismatch are unrelated documentation issues." }, { "left": "issue:43784", "right": "issue:43824", "accept": false, - "reason": "Different import errors for different missing classes." + "reason": "Importing sentence-transformers with a missing `nn` symbol is unrelated to a missing Qwen VL export." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-patch helper in GLM46V and InternVL2 meta-init incompatibility are different multimodal loading bugs." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "KV-cache continuation crash in serve is unrelated to the auto-docstring `__future__` annotations crash." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Different tokenizer regressions affecting different models and code paths." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Kimi-K2.5 codec handling and warning behavior is a different tokenizer regression than `clean_up_tokenization_spaces`." } ] }, - "evaluator_result": { - "accept": true, - "feedback": "The summary is grounded in the packet and the accepted duplicate pair is well supported by identical titles and matching bug description. The remaining soft-edge rejections are conservative and do not overstate similarity." - }, - "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 }, @@ -13077,19 +13011,19 @@ "issue:43475", "issue:43479", "issue:43723", - "issue:43874", "issue:43976", "issue:44060", "issue:44075", "issue:44079", "issue:44206", "issue:44222", + "issue:44263", "issue:44295", "issue:44297", "issue:44448", + "issue:44484", "issue:44623", "issue:44734", - "issue:44779", "issue:44869", "issue:44933", "issue:44977", @@ -13099,7 +13033,6 @@ "issue:45125", "issue:45245", "issue:45310", - "issue:45356", "issue:45381", "issue:45405", "issue:45656", @@ -13113,16 +13046,12 @@ "issue:46424" ], "soft_pairs": [ - "issue:43874|issue:45092", - "issue:44734|issue:45103", - "issue:42898|issue:44779", - "issue:42898|issue:45356", "issue:44295|issue:45656", "issue:43723|issue:46326", "issue:42175|issue:44933", "issue:42898|issue:44448", - "issue:45405|issue:45712", "issue:44060|issue:45125", + "issue:45405|issue:45712", "issue:44075|issue:46231", "issue:44222|issue:44297", "issue:43976|issue:44933", @@ -13146,240 +13075,230 @@ "issue:44206|issue:45103", "issue:43232|issue:46424", "issue:45381|issue:46305", - "issue:44734|issue:46421" + "issue:44734|issue:46421", + "issue:44263|issue:45092", + "issue:44222|issue:44484" ], - "prepared_review_unit_hash": "717b78e3244fb09d3f9b18fd45c4ef70e5f71bbe64e78f2ddc7cd9f021aadf77", + "prepared_review_unit_hash": "46b73aa037e7d70902bd388c8391c859dcd54e43e0eb09e2bd512d9362571596", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 34, - "serialized_chars": 37894, - "estimated_input_tokens": 9474, - "estimated_eval_tokens": 19204 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36608, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is noisy and heterogeneous: the issues span tokenizer regressions, multimodal model loading, cache/generation bugs, training/optimizer behavior, packaging/imports, and docs. I do not see any true duplicate pairs among the soft edges, so all candidate merges should be rejected.", - "confidence": 0.96, - "canonical_issue_reason": null, + "summary": "This is a loose set of Transformers issues with many lexical similarities but little evidence of true duplication. Most soft pairs are related only at a broad subsystem level (save_pretrained, cache, imports, generation, config) and should stay separate.", + "confidence": 0.77, + "canonical_issue_reason": "No strong canonical issue emerges: the items span unrelated bugs and regressions across model loading, tokenizers, training, cache, config, dependency, and docs paths, so there is no single clear duplicate anchor.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "If one issue must be used as a representative anchor, issue 44297 is the most concrete, user-facing serialization bug, but it still does not meaningfully subsume the rest of the set.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43874", - "right": "issue:45092", - "accept": false, - "reason": "Both involve vision/multimodal loading, but the failures are different: missing image-processor method vs remote-code checkpoint/meta-init incompatibility." - }, - { - "left": "issue:44734", - "right": "issue:45103", - "accept": false, - "reason": "Serve KV-cache continuation indexing crash is unrelated to the auto-docstring annotations AttributeError." - }, - { - "left": "issue:42898", - "right": "issue:44779", - "accept": false, - "reason": "Both are tokenizer regressions in v5, but one is clean_up_tokenization_spaces behavior and the other is Deepseek-specific incorrect tokenization/codec handling." - }, - { - "left": "issue:42898", - "right": "issue:45356", - "accept": false, - "reason": "Different tokenizer regressions: clean_up_tokenization_spaces vs Kimi-K2.5 codec handling and warning behavior." - }, { "left": "issue:44295", "right": "issue:45656", "accept": false, - "reason": "Position_ids buffer access bug and deepspeed double-optimizer-step bug are unrelated code paths." + "reason": "Different code paths: buffer registration/readout vs optimizer stepping under deepspeed. Same broad training/runtime area, not the same bug." }, { "left": "issue:43723", "right": "issue:46326", "accept": false, - "reason": "AutoTokenizer.from_pretrained loading issue is not the same as mistral_common continue_final_message behavior." + "reason": "Tokenizer loading regression vs mistral_common continue_final_message behavior. No shared underlying failure." }, { "left": "issue:42175", "right": "issue:44933", "accept": false, - "reason": "Torch-only install missing TensorFlow backend and a missing import from image_utils are separate packaging/import problems." + "reason": "Both are import/dependency complaints, but they concern different missing symbols/packages and unrelated fixes." }, { "left": "issue:42898", "right": "issue:44448", "accept": false, - "reason": "Tokenizer cleanup-space regression is not the same as Pegasus v4/v5 output divergence." + "reason": "One is a v5 tokenization cleanup behavior change; the other is a model output difference claim. Too indirect to treat as one bug." }, { - "left": "issue:45405", - "right": "issue:45712", + "left": "issue:44060", + "right": "issue:45125", "accept": false, - "reason": "PEFT version pinning issue and leftover dummy classes are unrelated maintenance problems." + "reason": "Different Qwen model bugs: a tied-weights warning issue versus missing tensor-parallel plan support." }, { - "left": "issue:44060", - "right": "issue:45125", + "left": "issue:45405", + "right": "issue:45712", "accept": false, - "reason": "Both concern Qwen models, but one is a tied-weights warning bug and the other is a missing tensor-parallel plan." + "reason": "Unrelated release/versioning problem vs dummy-class cleanup failure." }, { "left": "issue:44075", "right": "issue:46231", "accept": false, - "reason": "SGD arguments not being used is a runtime optimizer bug, not a docs parameter-name mismatch." + "reason": "Optimizer behavior bug vs documentation parameter-name mismatch." }, { "left": "issue:44222", "right": "issue:44297", "accept": false, - "reason": "FP8 save_pretrained/moe behavior and tokenizer_class serialization mismatch are different save-path bugs." + "reason": "Both mention save_pretrained, but one is FP8 MoE serialization and the other is tokenizer config metadata; not the same bug." }, { "left": "issue:43976", "right": "issue:44933", "accept": false, - "reason": "Python-version compatibility and image_utils import failure do not describe the same issue." + "reason": "Python version compatibility issue vs a missing image_utils import; unrelated causes and fixes." }, { "left": "issue:44933", "right": "issue:45712", "accept": false, - "reason": "Missing image_utils import and dummy_pt_objects leftovers are unrelated." + "reason": "Different import-related problems: a missing image_utils export vs leftover dummy classes in pt objects." }, { "left": "issue:44297", "right": "issue:45656", "accept": false, - "reason": "Tokenizer config serialization bug is unrelated to deepspeed optimizer stepping twice." + "reason": "Tokenizer serialization metadata vs optimizer double-step under deepspeed are unrelated." }, { "left": "issue:27957", "right": "issue:30333", "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and MLFlow reporting after failed training are unrelated." + "reason": "Flash Attention model support vs MLflow reporting on failed jobs; no common underlying defect." }, { "left": "issue:44222", "right": "issue:44623", "accept": false, - "reason": "FP8 MOE save_pretrained issue and processor.save_pretrained missing files are different save/export bugs." + "reason": "Both involve save_pretrained, but one is FP8 MoE saving and the other is processor file emission; distinct bugs." }, { "left": "issue:30333", "right": "issue:34689", "accept": false, - "reason": "MLFlow failed-job reporting is unrelated to Llama 3.2 Vision model loading regression." + "reason": "Training-job/MLflow reporting bug vs Llama 3.2 Vision model loading regression." }, { "left": "issue:44448", "right": "issue:46132", "accept": false, - "reason": "Pegasus output drift across versions and AttentionInterface.register changing function behavior are different bugs." + "reason": "Version-difference output regression vs AttentionInterface.register behavior change; too broad and not the same code path." }, { "left": "issue:42831", "right": "issue:44623", "accept": false, - "reason": "FineGrainedFP8 accuracy issue is unrelated to missing processor save artifacts." + "reason": "FP8 accuracy issue vs processor.save_pretrained file omission; unrelated symptoms." }, { "left": "issue:43479", "right": "issue:44079", "accept": false, - "reason": "Multimodal config default-init bug and ModelOutput None-key assignment bug affect different abstractions." + "reason": "Multimodal config initialization bug vs ModelOutput key assignment bug." }, { "left": "issue:30333", "right": "issue:36331", "accept": false, - "reason": "MLFlow reporting on failed jobs and CustomTrainer.compute_loss signature mismatch are unrelated." + "reason": "MLflow reporting failure vs Trainer compute_loss signature incompatibility." }, { "left": "issue:30333", "right": "issue:35141", "accept": false, - "reason": "MLFlow tracking after job failure is unrelated to resized embeddings being reinitialized." + "reason": "Failed-job reporting vs embedding reinitialization after resize; unrelated." }, { "left": "issue:45245", "right": "issue:46132", "accept": false, - "reason": "Category-count overflow and AttentionInterface.register behavior change are unrelated." + "reason": "Category cardinality limit error vs attention registration behavior change; no shared bug." }, { "left": "issue:43479", "right": "issue:44869", "accept": false, - "reason": "Multimodal config default construction and Whisper word timestamp decode crash are different issues." + "reason": "Config defaulting bug vs Whisper timestamp decode crash; unrelated." }, { "left": "issue:44297", "right": "issue:45310", "accept": false, - "reason": "Tokenizer_class serialization mismatch and Qwen3.5 MoE from_pretrained error are not the same bug." + "reason": "Tokenizer save_pretrained metadata mismatch vs Qwen3.5 MoE from_pretrained failure." }, { "left": "issue:41628", "right": "issue:42757", "accept": false, - "reason": "Missing AutoImageProcessor import and missing is_offline_mode import are separate symbol-export regressions." + "reason": "Different missing imports from different modules; same symptom class, not same defect." }, { "left": "issue:45071", "right": "issue:46082", "accept": false, - "reason": "PretrainedConfig type-checking regression and LlamaConfig head_dim validation are unrelated config bugs." + "reason": "Type-checking regression vs LlamaConfig head_dim validation; unrelated." }, { "left": "issue:43475", "right": "issue:46305", "accept": false, - "reason": "SAM3 missing attribute and MoonshineStreaming off-by-one masking are different model-specific failures." + "reason": "SAM 3 Video attribute error vs MoonshineStreaming mask off-by-one; different model families and code paths." }, { "left": "issue:43723", "right": "issue:44977", "accept": false, - "reason": "Tokenizer loading failure and Qwen3.5 flash-attention generation failure are unrelated." + "reason": "Tokenizer loading regression vs Qwen3.5 flash-attention generation bug." }, { "left": "issue:44206", "right": "issue:45103", "accept": false, - "reason": "LasrFeatureExtractor center-arg crash and auto-docstring annotations crash are different code paths." + "reason": "Unsupported feature-extractor arg vs auto_docstring crash with future annotations; unrelated." }, { "left": "issue:43232", "right": "issue:46424", "accept": false, - "reason": "Generation kwargs update after sync_gpus and max_cache_length handling are different cache/generation behaviors." + "reason": "Generation kwargs update bug vs cache max length constraint handling; both cache-related but not the same issue." }, { "left": "issue:45381", "right": "issue:46305", "accept": false, - "reason": "Qwen2.5-VL vision_position_ids bug and MoonshineStreaming sliding-window off-by-one are unrelated." + "reason": "Wrong vision_position_ids in Qwen2.5-VL video input vs MoonshineStreaming sliding-window off-by-one." }, { "left": "issue:44734", "right": "issue:46421", "accept": false, - "reason": "Serve KV-cache continuation indexing crash and lazy cache init causing recompilation are not the same failure." + "reason": "Serving KV-cache indexing crash vs lazy cache initialization causing recompilation; related domain, different bugs." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split indexing logic bug vs remote-code checkpoint incompatibility with v5 meta init; unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save_pretrained bug vs max_shard_size default question; both touch saving, but not the same change." } ] }, @@ -13410,6 +13329,7 @@ "issue:38175", "issue:39401", "issue:41628", + "issue:42757", "issue:42831", "issue:42907", "issue:43122", @@ -13423,14 +13343,12 @@ "issue:43873", "issue:44038", "issue:44190", - "issue:44222", "issue:44261", "issue:44263", "issue:44295", "issue:44360", "issue:44373", "issue:44483", - "issue:44484", "issue:44485", "issue:44568", "issue:44623", @@ -13439,6 +13357,7 @@ "issue:44913", "issue:44945", "issue:45092", + "issue:45278", "issue:45310", "issue:45357", "issue:45406", @@ -13453,8 +13372,6 @@ "issue:46431" ], "soft_pairs": [ - "issue:44263|issue:45092", - "issue:44222|issue:44484", "issue:43873|issue:44945", "issue:46209|issue:46431", "issue:44295|issue:44623", @@ -13482,72 +13399,62 @@ "issue:43408|issue:46402", "issue:22355|issue:27957", "issue:44945|issue:45753", - "issue:39401|issue:44568" + "issue:39401|issue:44568", + "issue:42757|issue:45278", + "issue:27957|issue:30990" ], - "prepared_review_unit_hash": "e77679d7a84990f95a081826c63d505d8cf44f41b6330fe443ebc8f3be1e22a6", + "prepared_review_unit_hash": "68265efbff3fe580f91868c266dd17b4de9eba88ae7c4276023b104f377d4bdf", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36315, - "estimated_input_tokens": 9079, - "estimated_eval_tokens": 18414 + "serialized_chars": 36321, + "estimated_input_tokens": 9081, + "estimated_eval_tokens": 18418 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: most items are unrelated bugs across different models, APIs, and subsystems. None of the soft pairs look like the same underlying issue, so all soft edges are rejected. The most central issue is the broad, discussion-heavy quantization/offloading report, but there is no strong true canonical duplicate hub here.", - "confidence": 0.91, - "canonical_issue_reason": "issue:43873 is the broadest and most discussion-heavy item in the set, but the cluster is too mixed to have a strong duplicate canonical; it is only the least-specific central candidate.", + "summary": "The cluster is heterogeneous: most items are distinct regression reports, model-specific bugs, or unrelated import/config issues. None of the soft pairs look like the same underlying bug/change, so I rejected all soft edges. The broadest representative issue is the transformers v5 import-regression report.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45278 is the broadest representative item in this set, covering upgrade-related import failures after moving to Transformers 5.5.0.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43873 is the best representative issue by breadth and activity, though it is not a clean canonical for the whole set because the items span many unrelated problems.", + "best_issue_reason": "issue:45278 is the best cluster representative because it is broad, regression-oriented, and closer to a catch-all umbrella than the more model-specific reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44263", - "right": "issue:45092", - "accept": false, - "reason": "Different bugs: torch.split return handling in a GLM MoE indexer vs checkpoint/meta-init incompatibility in old InternVL2." - }, - { - "left": "issue:44222", - "right": "issue:44484", - "accept": false, - "reason": "Both mention save_pretrained, but one is FP8 MoE saving and the other is max_shard_size behavior; not the same bug." - }, { "left": "issue:43873", "right": "issue:44945", "accept": false, - "reason": "Quantization/offloading behavior is unrelated to incorrect output under pipeline parallelism." + "reason": "Different failures: quantization/offloading behavior versus pipeline-parallel output correctness." }, { "left": "issue:46209", "right": "issue:46431", "accept": false, - "reason": "FP8 inference on B200 is unrelated to missing chat-template support in a processor." + "reason": "Unrelated subsystems: FP8 inference on B200 versus missing chat template in a processor." }, { "left": "issue:44295", "right": "issue:44623", "accept": false, - "reason": "Buffer registration/readback error is unrelated to processor.save_pretrained missing files." + "reason": "One is a buffer read/position_ids error; the other is processor.save_pretrained missing files." }, { "left": "issue:44373", "right": "issue:44485", "accept": false, - "reason": "A wrong docstring for position_ids is not the same as GLM-5 RoPE implementation behavior." + "reason": "Docstring issue versus GLM-5 RoPE implementation bug; not the same code-path problem." }, { "left": "issue:44190", @@ -13559,139 +13466,151 @@ "left": "issue:43232", "right": "issue:45834", "accept": false, - "reason": "Generation kwargs after sync_gpus and an OCR index error are different code paths and symptoms." + "reason": "Generation kwargs/sync_gpus behavior is unrelated to Kosmos2.5 OCR index errors." }, { "left": "issue:43122", "right": "issue:46209", "accept": false, - "reason": "Tokenizer behavior across versions is unrelated to an FP8 inference failure on B200." + "reason": "Tokenizer version drift is unrelated to FP8 inference on Hopper/B200." }, { "left": "issue:43723", "right": "issue:44871", "accept": false, - "reason": "Tokenizer loading failure and Gemma-3 eos_token_id mismatch are different model/config issues." + "reason": "Tokenizer loading regression and Gemma eos_token_id mismatch are different issues." }, { "left": "issue:44483", "right": "issue:45406", "accept": false, - "reason": "Chat/completions request rejection is unrelated to Gemma4Processor missing _tokenizer." + "reason": "Chat/completions request handling is not the same as Gemma4Processor missing _tokenizer." }, { "left": "issue:44263", "right": "issue:44360", "accept": false, - "reason": "Same broad DSA area, but torch.split return values vs missing ReLU are distinct bugs." + "reason": "Both mention DSA/GlmMoe code, but they describe different defects and code paths." }, { "left": "issue:43479", "right": "issue:43644", "accept": false, - "reason": "Phi4MultimodalConfig default-init behavior is unrelated to non-persistent buffer junk values." + "reason": "Default multimodal config initialization is unrelated to non-persistent buffer corruption on load." }, { "left": "issue:42907", "right": "issue:46050", "accept": false, - "reason": "Both involve saving quantized models, but they concern different quantization methods and failure modes." + "reason": "Different quantization/save-load failures on different models and mechanisms; not mergeable as one bug." }, { "left": "issue:30990", "right": "issue:38175", "accept": false, - "reason": "Sentence Transformers loading hangs and zero probabilities in SigLIP are unrelated." + "reason": "Sentence Transformers loading hang is unrelated to zero probabilities in SIGLIP2 classification." }, { "left": "issue:44261", "right": "issue:45092", "accept": false, - "reason": "Missing rms_norm_eps in MLA q_a_layernorm is unrelated to InternVL2 meta-initialization incompatibility." + "reason": "Missing rms_norm_eps precision issue and meta-initialization checkpoint incompatibility are distinct bugs." }, { "left": "issue:45092", "right": "issue:45357", "accept": false, - "reason": "Different models and operations: remote-code meta-init incompatibility vs incorrect visual-encoder keys on save_pretrained." + "reason": "Both are Qwen3.5-related, but one is checkpoint/meta-init incompatibility and the other is incorrect visual encoder keys on save." }, { "left": "issue:43644", "right": "issue:46305", "accept": false, - "reason": "Broken non-persistent buffer values are unrelated to a sliding-window mask off-by-one bug." + "reason": "Non-persistent buffer initialization bug is unrelated to MoonshineStreaming\u2019s off-by-one mask window." }, { "left": "issue:42831", "right": "issue:45310", "accept": false, - "reason": "FineGrainedFP8 accuracy regression is unrelated to Qwen3.5 MoE from_pretrained errors." + "reason": "FineGrainedFP8 accuracy problems are not the same as Qwen3.5 MoE from_pretrained errors." }, { "left": "issue:44829", "right": "issue:46002", "accept": false, - "reason": "Flash-attention-3 degenerate training and NaNs after checkout/init are different failure mechanisms." + "reason": "FlashAttention-3 degenerate training and NaNs after checkout/init are different training stability bugs." }, { "left": "issue:22355", "right": "issue:28282", "accept": false, - "reason": "Import errors for transformers.onnx and missing PyTorch are different dependency/module problems." + "reason": "Missing transformers.onnx module is unrelated to AutoModel failing because PyTorch is absent." }, { "left": "issue:44263", "right": "issue:44373", "accept": false, - "reason": "Torch.split return semantics and a docstring typo are not the same bug." + "reason": "A tensor split bug and a docstring bug are not the same underlying issue." }, { "left": "issue:42831", "right": "issue:44038", "accept": false, - "reason": "FineGrainedFP8 accuracy and Qwen3-VL-Moe regression are separate issues." + "reason": "Different problems: FP8 accuracy versus a Qwen3-VL-Moe regression in Transformers 5." }, { "left": "issue:41628", "right": "issue:43824", "accept": false, - "reason": "These are different missing symbols; they may both be import regressions, but not the same concrete code-path bug." + "reason": "Import errors for different symbols are both broad import regressions, but they do not point to the same broken code path." }, { "left": "issue:43479", "right": "issue:44913", "accept": false, - "reason": "Default config initialization and rotary_pct reloading behavior are unrelated." + "reason": "Incorrect default multimodal config initialization is unrelated to GPTNeoX rotary_pct reload behavior." }, { "left": "issue:43761", "right": "issue:46082", "accept": false, - "reason": "CLIPVisionModel hidden_states regression is unrelated to LlamaConfig head_dim validation." + "reason": "CLIP vision hidden_states regression and Llama head_dim validation are unrelated model-specific issues." }, { "left": "issue:43408", "right": "issue:46402", "accept": false, - "reason": "Sam3Model type warning and missing text-encoder weights are distinct SAM3 issues." + "reason": "Sam3 tracker/model-type warning is unrelated to text encoder weights not being loaded." }, { "left": "issue:22355", "right": "issue:27957", "accept": false, - "reason": "Transformers.onnx import failure is unrelated to XLMRoberta Flash Attention 2 support." + "reason": "Missing transformers.onnx and XLMRoberta Flash Attention 2 are distinct issues." }, { "left": "issue:44945", "right": "issue:45753", "accept": false, - "reason": "Pipeline-parallelism output errors and Qwen3_5 infinite-loop behavior are different bugs." + "reason": "Pipeline-parallel output correctness and Qwen3_5 image loop are different model/runtime bugs." }, { "left": "issue:39401", "right": "issue:44568", "accept": false, - "reason": "Qwen3 offset_mapping is a tokenizer bug, while mdeberta add_special_tokens missing BOS/EOS is a different tokenizer behavior change." + "reason": "Tokenizer offset_mapping bug is unrelated to add_special_tokens not adding BOS/EOS." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing dependency symbol and a broad multi-import regression are related in theme only, not the same bug." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 on XLMRoberta is unrelated to Sentence Transformers hanging during load." } ] }, @@ -13715,12 +13634,10 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:27957", "issue:30990", "issue:33357", "issue:41628", "issue:42371", - "issue:42757", "issue:42796", "issue:42886", "issue:42907", @@ -13730,6 +13647,7 @@ "issue:43122", "issue:43232", "issue:43408", + "issue:43452", "issue:43475", "issue:43550", "issue:43576", @@ -13749,6 +13667,7 @@ "issue:44368", "issue:44373", "issue:44485", + "issue:44514", "issue:44625", "issue:44704", "issue:44829", @@ -13765,8 +13684,6 @@ "issue:46369" ], "soft_pairs": [ - "issue:42757|issue:45278", - "issue:27957|issue:30990", "issue:42886|issue:43066", "issue:44368|issue:46305", "issue:42796|issue:43723", @@ -13794,216 +13711,218 @@ "issue:44038|issue:44295", "issue:43756|issue:44060", "issue:44625|issue:44829", - "issue:43408|issue:43475" + "issue:43408|issue:43475", + "issue:42907|issue:43452", + "issue:43122|issue:44514" ], - "prepared_review_unit_hash": "ede581e9c7ba9c4fd0f4421d4265b423b9309e50c5efa786295debf7d3f879a0", + "prepared_review_unit_hash": "3f4ccd3fc2c54a535ab646f50675c864ee6dcd98d56ecc5f88e4e8e8f36282e0", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 30, - "serialized_chars": 36319, - "estimated_input_tokens": 9080, - "estimated_eval_tokens": 18416 + "serialized_chars": 36419, + "estimated_input_tokens": 9105, + "estimated_eval_tokens": 18466 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a grab-bag of distinct Transformers regressions (imports, tokenizers, generation, model-specific bugs). The closest thing to an umbrella issue is a broad v5 import-regression report, but the soft pairs do not look like true duplicates of one another.", - "confidence": 0.81, - "canonical_issue_reason": "issue:45278 is the broadest, most general user-facing regression report in the set: multiple import errors after upgrading to 5.5.0. It is the best umbrella issue for this mixed cluster, even though most other items are not duplicates of it.", + "summary": "Mostly unrelated issue reports, with a loose theme of Transformers v5 regressions around imports, tokenizers, and model loading. None of the soft-edge pairs look like true same-bug duplicates.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: a general upgrade-to-v5 import regression that can reasonably represent the cluster better than the narrower, model-specific bugs.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45278 best represents the cluster because it captures the broad v5 upgrade/import breakage theme and is more central than the narrower model- or tokenizer-specific reports.", + "best_issue_reason": "issue:45278 is the most representative user-facing failure in the set and the best anchor if one issue must stand in for the cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42757", - "right": "issue:45278", - "accept": false, - "reason": "Both mention import failures, but one is a specific huggingface_hub symbol and the other is a broad post-upgrade import regression; not the same underlying bug." - }, - { - "left": "issue:27957", - "right": "issue:30990", - "accept": false, - "reason": "XLMRoberta Flash Attention 2 support and Sentence Transformers hanging on load are unrelated code paths and symptoms." - }, { "left": "issue:42886", "right": "issue:43066", "accept": false, - "reason": "Offline cache loading failure and wrong tokenizer decoder type are different tokenizer-related bugs." + "reason": "Offline cache loading failure vs wrong decoder-type mapping; related tokenizer area, but different concrete bugs." }, { "left": "issue:44368", "right": "issue:46305", "accept": false, - "reason": "A tied-embeddings warning in Qwen3.5 fine-tuning is unrelated to a MoonshineStreaming sliding-window off-by-one bug." + "reason": "Both mention warnings/behavior in model internals, but one is a tied-weights warning and the other is a sliding-window off-by-one." }, { "left": "issue:42796", "right": "issue:43723", "accept": false, - "reason": "Both involve tokenizers, but one is a BPE marker handling bug and the other is a tokenizer loading regression; not the same defect." + "reason": "Specific tokenizer BPE-marker handling bug vs generic AutoTokenizer loading failure; not the same defect." }, { "left": "issue:42886", "right": "issue:42994", "accept": false, - "reason": "Cache/offline loading and quantized model saving are separate failure modes." + "reason": "Tokenizer offline-cache loading and quantized-model saving are unrelated failure modes." }, { "left": "issue:43408", "right": "issue:46369", "accept": false, - "reason": "A SAM3 model-type mismatch warning is unrelated to a memory overflow in sam3-litetext." + "reason": "Same SAM3 family, but a type-mismatch warning is not the same as a memory overflow." }, { "left": "issue:43550", "right": "issue:46132", "accept": false, - "reason": "torch.compile with SDPA on Bamba and AttentionInterface.register behavior changes are different implementation bugs." + "reason": "One is a Bamba torch.compile/SDPA bug; the other is an AttentionInterface registration behavior change." }, { "left": "issue:43646", "right": "issue:46002", "accept": false, - "reason": "Custom model initialization breakage and NaNs in classification heads are not the same code-path problem." + "reason": "Custom model initialization breakage and NaNs in classification heads are different code paths and symptoms." }, { "left": "issue:43576", "right": "issue:43784", "accept": false, - "reason": "The env CLI regression and the sentence-transformers import NameError are separate issues." + "reason": "Broken env command vs import-time NameError in sentence-transformers; not the same bug." }, { "left": "issue:44295", "right": "issue:44297", "accept": false, - "reason": "Reading position_ids after buffer registration and tokenizer_class mismatch on save are unrelated." + "reason": "Position_ids buffer access error and tokenizer_class mismatch on save are separate tokenizer/model save issues." }, { "left": "issue:42371", "right": "issue:44263", "accept": false, - "reason": "TF32 API settings guidance and a torch.split return-value issue do not share the same underlying bug." + "reason": "TF32 settings warning is unrelated to the DSA indexer torch.split issue." }, { "left": "issue:44117", "right": "issue:44704", "accept": false, - "reason": "TOKENIZER_MAPPING_NAMES returning None and AutoProcessor not forwarding kwargs are different loader bugs." + "reason": "Mapping-lookup None handling and cached_file kwarg forwarding are different API bugs." }, { "left": "issue:30990", "right": "issue:33357", "accept": false, - "reason": "Sentence Transformers load hang and a MacOS bus error with a CLIP model are different failure modes." + "reason": "SentenceTransformers loading hang and Mac bus error on CLIP are distinct runtime failures." }, { "left": "issue:42907", "right": "issue:44945", "accept": false, - "reason": "Quantized model save failure and incorrect pipeline-parallel outputs are unrelated." + "reason": "Quantized-model save failure and pipeline-parallel output corruption are unrelated." }, { "left": "issue:43232", "right": "issue:45593", "accept": false, - "reason": "Generation kwargs update logic and missing auxiliary losses in D-FINE are separate problems." + "reason": "Generation kwargs after sync_gpus and D-FINE auxiliary-loss omission do not share a code path." }, { "left": "issue:44261", "right": "issue:44373", "accept": false, - "reason": "A precision/config bug in MLA layernorm and a wrong docstring are not duplicates." + "reason": "MLA precision/config issue and a wrong docstring are not duplicate bugs." }, { "left": "issue:42796", "right": "issue:43122", "accept": false, - "reason": "Tokenizer BPE marker handling and a tokenization difference across versions are similar only at a high level, not the same concrete bug." + "reason": "Both involve tokenization, but one is a model-specific BPE marker bug and the other is a version-to-version tokenization change." }, { "left": "issue:42907", "right": "issue:43937", "accept": false, - "reason": "Saving dequantized models and an invalid GenerationConfig error are unrelated." + "reason": "Saving dequantized models and invalid GenerationConfig are unrelated." }, { "left": "issue:41628", "right": "issue:45278", "accept": false, - "reason": "Both are import-related regressions, but one is a specific missing AutoImageProcessor symbol while the other is a broad upgrade-wide import breakage; not the same artifact." + "reason": "Missing AutoImageProcessor import is one instance of v5 import breakage, but 45278 is a broad umbrella rather than the same concrete bug." }, { "left": "issue:44704", "right": "issue:45290", "accept": false, - "reason": "Passing kwargs to cached_file and chat-template crashes with tool calls are different code paths." + "reason": "AutoProcessor kwarg forwarding and chat-template crashes with tool-call messages are different processor bugs." }, { "left": "issue:42907", "right": "issue:44964", "accept": false, - "reason": "Quantized model saving and loading Phi-4 multimodal instruct are different issues." + "reason": "Quantized save failure and loading Phi-4 multimodal are unrelated model-loading problems." }, { "left": "issue:43784", "right": "issue:45600", "accept": false, - "reason": "Sentence-transformers import NameError and auto_mappings references to removed configs are separate import/CI breakages." + "reason": "Sentence-transformers import NameError and stale auto_mappings references are different regression types." }, { "left": "issue:44038", "right": "issue:44297", "accept": false, - "reason": "A Qwen3-VL-Moe bug and a tokenizer save metadata mismatch are unrelated." + "reason": "Qwen3-VL-Moe bug and tokenizer save metadata mismatch are not the same issue." }, { "left": "issue:44360", "right": "issue:44485", "accept": false, - "reason": "A missing ReLU in DSA indexer and GLM-5 RoPE implementation are different model internals." + "reason": "DSA indexer activation omission and GLM-5 RoPE implementation are unrelated architecture bugs." }, { "left": "issue:43116", "right": "issue:45200", "accept": false, - "reason": "Example-script multi-label empty outputs and Gemma 4 mm_token_type_ids defaults are unrelated classification/model-input issues." + "reason": "Example-script multi-label output bug and Gemma 4 text-only token-type defaults are different behaviors." }, { "left": "issue:44038", "right": "issue:44295", "accept": false, - "reason": "Qwen3-VL-Moe behavior and position_ids buffer access are different bugs." + "reason": "Qwen3-VL-Moe issue and position_ids buffer access error have no shared failure mechanism." }, { "left": "issue:43756", "right": "issue:44060", "accept": false, - "reason": "Smollm3 RoPE-layer dropping and Qwen3-Next tied-weights warnings are unrelated model-specific regressions." + "reason": "Smollm3 RoPE layer mismatch and Qwen3-Next tied-weights warning are separate model-specific config problems." }, { "left": "issue:44625", "right": "issue:44829", "accept": false, - "reason": "num_labels propagation and flash_attention_3 training degeneration are distinct failures." + "reason": "num_labels propagation and flash_attention_3 training collapse are unrelated bugs." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Both are SAM 3 Video related, but model-type warning and missing fpn_position_embeddings are different defects." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Quantized-model saving and gguf_file loading failures are distinct pre/post-save code paths." }, { - "left": "issue:43408", - "right": "issue:43475", + "left": "issue:43122", + "right": "issue:44514", "accept": false, - "reason": "The model-type mismatch warning and missing fpn_position_embeddings attribute are both SAM3-related but not the same concrete bug." + "reason": "Tokenizer output differences across versions and batched chat-template crashes are not the same bug." } ] }, @@ -14033,16 +13952,15 @@ "issue:29159", "issue:36296", "issue:39401", + "issue:42175", "issue:42222", "issue:42371", "issue:42757", "issue:42831", - "issue:42907", "issue:43066", "issue:43116", "issue:43122", "issue:43441", - "issue:43452", "issue:43493", "issue:43611", "issue:43824", @@ -14060,11 +13978,11 @@ "issue:44355", "issue:44485", "issue:44488", - "issue:44514", "issue:44521", "issue:44704", "issue:44779", "issue:44936", + "issue:44938", "issue:45083", "issue:45216", "issue:45357", @@ -14077,8 +13995,6 @@ "issue:46328" ], "soft_pairs": [ - "issue:42907|issue:43452", - "issue:43122|issue:44514", "issue:44038|issue:44222", "issue:43122|issue:44779", "issue:44279|issue:46328", @@ -14088,8 +14004,8 @@ "issue:39401|issue:44488", "issue:39401|issue:44521", "issue:22355|issue:28218", - "issue:22355|issue:29159", "issue:42757|issue:44351", + "issue:22355|issue:29159", "issue:36296|issue:43116", "issue:43992|issue:44704", "issue:43066|issue:44779", @@ -14108,228 +14024,223 @@ "issue:43116|issue:43493", "issue:45083|issue:45446", "issue:22355|issue:29127", - "issue:43937|issue:45563" + "issue:43937|issue:45563", + "issue:42175|issue:44938" ], - "prepared_review_unit_hash": "a975d3cfd060252fffe1935eb5fff8141b727e9a88c5c43c80bd7bb385879da8", + "prepared_review_unit_hash": "7f9d60c97c7a6911934bb92b852cb77de5c4f2823479ffa6411b2d4984262046", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 32, - "serialized_chars": 36728, - "estimated_input_tokens": 9182, - "estimated_eval_tokens": 18620 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 35813, + "estimated_input_tokens": 8954, + "estimated_eval_tokens": 18164 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: it mixes tokenizer regressions, model loading/saving bugs, import failures, and unrelated performance/accuracy issues. There is no clear duplicate pair strong enough to merge, and no PRs are present.", - "confidence": 0.58, - "canonical_issue_reason": "Issue 43122 is the broadest and most central tokenizer-regression report in the set, with the most overlap to other tokenizer-related issues.", + "summary": "The cluster is not a true duplicate set; it mixes unrelated tokenizer regressions, model-specific load/save bugs, import errors, and documentation/performance issues. No soft edge looks like the same underlying bug/change with enough precision to merge.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43122 is the broadest and most representative tokenizer-regression report in the set, so it is the least-bad anchor for the small tokenizer-related subgroup, though the overall cluster is still heterogeneous.", "canonical_pr_reason": null, - "best_issue_reason": "Issue 43122 is the best umbrella candidate for the tokenizer subcluster because it is broad, open, and closest to several related tokenization reports.", + "best_issue_reason": "issue:43122 is the best single representative issue because it has the clearest cross-version tokenizer regression framing and the strongest chance of being an umbrella for nearby tokenizer complaints; it still does not justify merging the whole cluster.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42907", - "right": "issue:43452", - "accept": false, - "reason": "Both involve model/file-loading regressions, but they describe different failure modes and code paths (dequantized save vs AutoTokenizer/AutoModel loading)." - }, - { - "left": "issue:43122", - "right": "issue:44514", - "accept": false, - "reason": "Both are tokenizer-related, but one is a version-to-version tokenization change while the other is a batched apply_chat_template crash; different bugs." - }, { "left": "issue:44038", "right": "issue:44222", "accept": false, - "reason": "These both mention MoE/FP8 areas, but one is a Qwen3-VL-Moe runtime bug and the other is FP8 save_pretrained behavior; not the same fix." + "reason": "Both mention moe models, but one is a Qwen3-VL-Moe load/inference bug and the other is FP8 save_pretrained behavior; different code paths and symptoms." }, { "left": "issue:43122", "right": "issue:44779", "accept": false, - "reason": "Both are tokenization regressions in v5, but they concern different models and symptoms (generic tokenization changes vs DeepSeek incorrect results)." + "reason": "Both are tokenizer regressions in v5, but one is a generic cross-version tokenization change and the other is Deepseek-specific incorrect output; not the same concrete bug." }, { "left": "issue:44279", "right": "issue:46328", "accept": false, - "reason": "Completely different problems: dependency resolution versus SequenceFeatureExtractor padding performance." + "reason": "A dependency complaint is unrelated to a SequenceFeatureExtractor.pad performance issue." }, { "left": "issue:43122", "right": "issue:44521", "accept": false, - "reason": "Tokenization overlap is too broad; assistant mask generation for multimodal inputs is a different code path and bug." + "reason": "Tokenization regression vs apply_chat_template assistant mask bug; different APIs and failure modes." }, { "left": "issue:43824", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but they are for different symbols and likely unrelated breakages." + "reason": "Both are import errors, but for different symbols and modules; no shared underlying fix is evident." }, { "left": "issue:45216", "right": "issue:45478", "accept": false, - "reason": "Both involve Qwen3.5 in v5, but one is incorrect checkpoint saving and the other is a from_pretrained loading error; distinct regressions." + "reason": "Both are Qwen3.5-related, but one is save_pretrained checkpoint corruption and the other is from_pretrained loading failure; too different to merge." }, { "left": "issue:39401", "right": "issue:44488", "accept": false, - "reason": "Different model issues: wrong offset mapping for Qwen3 tokenizer versus failure to load cjvt/sleng-bert." + "reason": "Qwen3 tokenizer offset_mapping bug vs loading failure for an unrelated model." }, { "left": "issue:39401", "right": "issue:44521", "accept": false, - "reason": "Qwen3 tokenizer offset mapping and multimodal assistant_mask behavior are unrelated bugs." + "reason": "Tokenizer offset mapping issue does not match a multimodal assistant_mask bug." }, { "left": "issue:22355", "right": "issue:28218", "accept": false, - "reason": "No module transformers.onnx and extra-space tokenization after added token are unrelated issues." + "reason": "ONNX import failure and added-token spacing are unrelated." }, { - "left": "issue:22355", - "right": "issue:29159", + "left": "issue:42757", + "right": "issue:44351", "accept": false, - "reason": "The ONNX import problem and slow/fast tokenizer inconsistency are different failure classes." + "reason": "Import errors, but from different packages and missing names; not the same defect." }, { - "left": "issue:42757", - "right": "issue:44351", + "left": "issue:22355", + "right": "issue:29159", "accept": false, - "reason": "Both are import-related, but they target different missing names from different packages/APIs." + "reason": "No module import problem vs slow/fast tokenizer inconsistency." }, { "left": "issue:36296", "right": "issue:43116", "accept": false, - "reason": "Tensor parallel training bug and multi-label classification script returning empty results are unrelated." + "reason": "Tensor parallel training bug and example-script classification output bug are unrelated." }, { "left": "issue:43992", "right": "issue:44704", "accept": false, - "reason": "Missing embed_tokens.weight in UMT5Encoder and cached_file kwargs forwarding are separate code paths." + "reason": "Both involve from_pretrained, but one misses a weight during model load while the other drops kwargs in processor file resolution." }, { "left": "issue:43066", "right": "issue:44779", "accept": false, - "reason": "Wrong tokenizer decoder type and DeepSeek tokenization regression are different tokenizer bugs." + "reason": "Wrong tokenizer decoder type serialization is different from Deepseek tokenization producing incorrect results." }, { "left": "issue:43937", "right": "issue:44485", "accept": false, - "reason": "GenerationConfig validation and RoPE implementation are not the same underlying issue." + "reason": "GenerationConfig validation error and GLM-5 RoPE implementation are distinct model issues." }, { "left": "issue:44355", "right": "issue:46231", "accept": false, - "reason": "Compiled Python file errors and a documentation parameter-name mismatch are unrelated." + "reason": "Running compiled Python files and a documentation parameter-name typo are unrelated." }, { "left": "issue:43611", "right": "issue:46002", "accept": false, - "reason": "Base_model_prefix loading regression and NaNs in classification heads are different bugs." + "reason": "Base model prefix loading regression and NaNs in classification heads are different bugs." }, { "left": "issue:42371", "right": "issue:43874", "accept": false, - "reason": "TF32 behavior guidance and a missing image-processor method are unrelated." + "reason": "TF32 behavior docs/API issue vs missing image processor method; no overlap." }, { "left": "issue:42831", "right": "issue:44297", "accept": false, - "reason": "FP8 accuracy issues and tokenizer_class metadata mismatch affect different subsystems." + "reason": "FP8 accuracy issue and tokenizer_class mismatch are separate save/load concerns." }, { "left": "issue:42222", "right": "issue:43441", "accept": false, - "reason": "VitPose breakage and Ministral FlashAttention failure are different model-specific regressions." + "reason": "VitPose model breakage and Ministral-3 FlashAttention failure are unrelated model-specific regressions." }, { "left": "issue:44038", "right": "issue:45478", "accept": false, - "reason": "Both mention Qwen3.5/MoE, but one is a Qwen3-VL-Moe bug and the other is a from_pretrained failure; not the same concrete problem." + "reason": "Both are Qwen/MoE related, but one is Qwen3-VL-Moe and the other is Qwen3.5 Moe from_pretrained; not the same concrete bug." }, { "left": "issue:43976", "right": "issue:45446", "accept": false, - "reason": "Python version support breakage and a PyTorch version-check bug are unrelated." + "reason": "Python version compatibility failure and a PyTorch version-check bug are unrelated." }, { "left": "issue:45357", "right": "issue:45478", "accept": false, - "reason": "Both involve Qwen3.5, but one is incorrect visual encoder keys on save and the other is a from_pretrained loading error." + "reason": "Both concern Qwen3.5 and version 5.x, but one is incorrect visual-encoder save_pretrained keys while the other is a from_pretrained load error." }, { "left": "issue:43493", "right": "issue:45753", "accept": false, - "reason": "SigLIP2 implementation discrepancy and Qwen3_5 infinite-loop behavior are different model regressions." + "reason": "SigLIP2 implementation discrepancy and Qwen3_5 image infinite loop are unrelated model issues." }, { "left": "issue:44075", "right": "issue:44936", "accept": false, - "reason": "SGD argument handling and trainer.evaluate() after train() are separate trainer/optimizer problems." + "reason": "Optimizer argument handling and trainer evaluate/train interaction are separate training bugs." }, { "left": "issue:43116", "right": "issue:43825", "accept": false, - "reason": "A multi-label classification script bug and a pipeline translation-task error message are unrelated." + "reason": "Example multi-label classification returning empty results is unrelated to a pipeline error message about translation support." }, { "left": "issue:43116", "right": "issue:43493", "accept": false, - "reason": "Example-script classification behavior and SigLIP2 implementation discrepancy are unrelated." + "reason": "Classification example script bug and SigLIP2 implementation discrepancy are unrelated." }, { "left": "issue:45083", "right": "issue:45446", "accept": false, - "reason": "Feature-extractor length calculation and AuxRequest import version checking are different concerns." + "reason": "qwen3_omni_moe feature-length helper behavior and flex_attention version gating are different code paths." }, { "left": "issue:22355", "right": "issue:29127", "accept": false, - "reason": "ONNX import failure and LayoutLMv3 box-clarity messaging are unrelated." + "reason": "Importing transformers.onnx and a LayoutLMv3 error-message clarity issue are unrelated." }, { "left": "issue:43937", "right": "issue:45563", "accept": false, - "reason": "GenerationConfig validation and stale paged-generate warning are different issues." + "reason": "GLM generation config validation and a stale warning in paged generate() are different generation bugs." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Backend packaging issue for pip install '.[torch]' and Python 3.14 load failure are unrelated environment problems." } ] }, @@ -14357,7 +14268,6 @@ "issue:38617", "issue:39401", "issue:41628", - "issue:42175", "issue:42222", "issue:42673", "issue:43064", @@ -14383,8 +14293,8 @@ "issue:44779", "issue:44829", "issue:44877", - "issue:44938", "issue:44945", + "issue:44964", "issue:44977", "issue:45071", "issue:45072", @@ -14394,6 +14304,7 @@ "issue:45310", "issue:45405", "issue:45446", + "issue:45478", "issue:45588", "issue:46002", "issue:46077", @@ -14403,7 +14314,6 @@ "issue:46328" ], "soft_pairs": [ - "issue:42175|issue:44938", "issue:38617|issue:45278", "issue:33453|issue:39401", "issue:44625|issue:46002", @@ -14431,216 +14341,227 @@ "issue:39401|issue:43421", "issue:43937|issue:44496", "issue:44117|issue:46328", - "issue:43066|issue:43122" + "issue:43066|issue:43122", + "issue:44964|issue:45278", + "issue:44410|issue:45478" ], - "prepared_review_unit_hash": "51cdc8bc87d4a13d9c3785ae19aad7ce8fdcb93e3369cc74f3f95de4d996b734", + "prepared_review_unit_hash": "84ace9a682172bce6794b58ea46dd9d0c992a5919cbc31d42329591f243939cc", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 29, - "serialized_chars": 36371, - "estimated_input_tokens": 9093, - "estimated_eval_tokens": 18442 + "soft_pair_count": 30, + "serialized_chars": 36702, + "estimated_input_tokens": 9176, + "estimated_eval_tokens": 18608 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a loose topical grouping of Transformers bug reports, not a true duplicate set. The only common thread is a broad v5/tokenizer/import-regression theme, but the candidate pairs describe different concrete failures and code paths. No PRs are present.", - "confidence": 0.83, - "canonical_issue_reason": "issue:43122 is the broadest, most representative issue here: it describes a version-wide tokenizer regression between 4.57.3 and 5.0, which best matches the cluster\u2019s dominant theme more than the narrower model-specific reports.", + "summary": "The cluster remains heterogeneous: it mixes distinct v5 regressions across tokenizers, model loading, attention/config handling, and docs. issue:45278 is still the best broad representative, but it is an umbrella rather than a true duplicate canonical.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set, covering multiple import failures after upgrading to v5.5.0, so it best represents the cluster even though the items are not one duplicate bug.", "canonical_pr_reason": null, - "best_issue_reason": "issue:43122 is the best single representative for this cluster because it captures the general tokenizer-regression pattern without being tied to one model or a niche failure mode.", + "best_issue_reason": "issue:45278 is the most generally applicable issue to represent this mixed cluster, but it should be treated as a coarse umbrella, not a precise duplicate target.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42175", - "right": "issue:44938", - "accept": false, - "reason": "Different failure modes: backend/extra dependency packaging vs Python 3.14 import compatibility. Same ecosystem, not the same bug." - }, { "left": "issue:38617", "right": "issue:45278", "accept": false, - "reason": "45278 is a broad umbrella of import errors, but this pair does not identify the same concrete missing symbol or code path." + "reason": "Both are import-related, but one is a specific missing symbol and the other is a broad upgrade regression; not the same concrete bug." }, { "left": "issue:33453", "right": "issue:39401", "accept": false, - "reason": "Both are tokenizer-related, but one is loading regression and the other is Qwen3 offset_mapping correctness; different symptoms and likely different fixes." + "reason": "Tokenizer loading regression vs. Qwen3 offset_mapping bug are different symptoms and likely different code paths." }, { "left": "issue:44625", "right": "issue:46002", "accept": false, - "reason": "Config propagation for Qwen3.5 num_labels is unrelated to NaNs in classification heads during checkout/init." + "reason": "num_labels propagation and NaNs in classification heads are unrelated model-behavior issues." }, { "left": "issue:43381", "right": "issue:44945", "accept": false, - "reason": "Gradient checkpointing eval-mode behavior is unrelated to incorrect outputs under pipeline parallelism." + "reason": "Eval-mode gradient checkpointing failure and pipeline-parallel wrong outputs are distinct training/runtime problems." }, { "left": "issue:44117", "right": "issue:45588", "accept": false, - "reason": "Tokenizer mapping assumption bug vs flash_attention sink-less-model AttributeError; no shared code path." + "reason": "Tokenizer mapping fallback and flash-attention crash on s_aux=None affect different subsystems." }, { "left": "issue:44380", "right": "issue:46077", "accept": false, - "reason": "Different attention bugs in different models/backends: GPT2 scaling ignored vs Gemma4 mask construction." + "reason": "Both concern attention behavior, but they involve different models and different mechanisms." }, { "left": "issue:39401", "right": "issue:44779", "accept": false, - "reason": "Both involve tokenizer output correctness, but they are model-specific regressions with different observed behavior (offset_mapping vs tokenization results)." + "reason": "Both are tokenizer regressions, but they are model-specific incorrect-tokenization reports without evidence of a shared root cause." }, { "left": "issue:43827", "right": "issue:45446", "accept": false, - "reason": "Docs still referencing pipeline() is unrelated to a PyTorch version check bug in flex_attention." + "reason": "Docs still referencing pipeline() and a PyTorch version-check bug are unrelated." }, { "left": "issue:43450", "right": "issue:43986", "accept": false, - "reason": "Batched video shape bug and missing-torchvision AutoProcessor crash are both in video processing, but not the same underlying issue." + "reason": "Batched video shape bug and missing-torchvision crash are different video-processing failures." }, { "left": "issue:43064", "right": "issue:43673", "accept": false, - "reason": "Trainer optimizer-state corruption under FSDP2/PEFT is unrelated to GenerationMixin cache missing during chunked_prefill." + "reason": "Optimizer-state corruption in training and missing generation cache during chunked prefill are unrelated." }, { "left": "issue:44466", "right": "issue:46002", "accept": false, - "reason": "Serialization of lm_head.weight depends on device; NaNs in classification heads are a different numerical/training issue." + "reason": "Serialization of tied weights and NaNs in classification heads are different failure modes." }, { "left": "issue:43937", "right": "issue:44977", "accept": false, - "reason": "GLM-5 GenerationConfig validation and Qwen3.5 flash-attention generation failure are different model/setup bugs." + "reason": "Invalid GenerationConfig for GLM-5 and Qwen3.5 flash-attention generation failure are different issues." }, { "left": "issue:45405", "right": "issue:45446", "accept": false, - "reason": "A bumped unreleased PEFT minimum version is unrelated to an incorrect PyTorch version gate for flex_attention." + "reason": "A released-version dependency bump and a PyTorch version check bug are not the same bug." }, { "left": "issue:44410", "right": "issue:45310", "accept": false, - "reason": "Both mention Qwen-family models, but one is missing projections in qwen3next layer 0 and the other is a from_pretrained error for Qwen3.5 MoE." + "reason": "Missing qkv/gate projections in qwen3next and a Qwen3.5 MoE from_pretrained error are different model-loading defects." }, { "left": "issue:43122", "right": "issue:46298", "accept": false, - "reason": "A tokenizer version-regression bug and a noisy logging message are not the same underlying issue." + "reason": "Tokenization regression and noisy logging are unrelated." }, { "left": "issue:38617", "right": "issue:41628", "accept": false, - "reason": "Both are import errors, but they concern different missing names and likely separate fixes." + "reason": "Both are import errors, but they involve different missing symbols and likely different regression causes." }, { "left": "issue:44117", "right": "issue:45081", "accept": false, - "reason": "Token mapping None handling is unrelated to the Mistral regex patch crashing on backend_tokenizer access." + "reason": "Tokenizer mapping None-handling and Mistral regex patch crash are different tokenizer-loading bugs." }, { "left": "issue:42673", "right": "issue:45072", "accept": false, - "reason": "VRAM leak in multi-threaded dataloader use is unrelated to bfloat16 dtype mismatches in inference." + "reason": "VRAM leak in Qwen3ForCausalLM threading and bfloat16 dtype mismatch in CI are unrelated." }, { "left": "issue:42222", "right": "issue:46178", "accept": false, - "reason": "VitPose model breakage and an undefined tensorflow_text backend are distinct loading/back-end issues." + "reason": "Broken vitpose models and an unsupported backend error are unrelated." }, { "left": "issue:44291", "right": "issue:45071", "accept": false, - "reason": "Unexpected init_empty_weights kwarg handling is different from PretrainedConfig type-checking breakage." + "reason": "Unexpected init_empty_weights argument and PretrainedConfig type-checking breakage are both init-path regressions, but not the same concrete bug." }, { "left": "issue:44829", "right": "issue:45216", "accept": false, - "reason": "FlashAttention-3 training degeneration and Qwen3.5 checkpoint serialization are different bugs with different lifecycles." + "reason": "FlashAttention-3 degenerate training and Qwen3.5 save_pretrained regression are unrelated." }, { "left": "issue:43408", "right": "issue:43742", "accept": false, - "reason": "Wrong model-type warning for sam3_tracker is unrelated to a MobileLLM key error on load." + "reason": "Wrong model-type warning and MobileLLM key error are different model-loading issues." }, { "left": "issue:44877", "right": "issue:46231", "accept": false, - "reason": "Strict config loading for granite_speech has nothing to do with a documentation parameter-name mismatch." + "reason": "Strict config loading and a documentation parameter-name typo are unrelated." }, { "left": "issue:44484", "right": "issue:45310", "accept": false, - "reason": "save_pretrained shard-size policy and Qwen3.5 from_pretrained failure are unrelated." + "reason": "save_pretrained shard-size behavior and Qwen3.5 from_pretrained failure are different bugs." }, { "left": "issue:39401", "right": "issue:43421", "accept": false, - "reason": "Both are tokenizer bugs, but one is offset_mapping on Qwen3 and the other is runtime post-processor updates after special-token edits; different code paths." + "reason": "Qwen3 offset mapping and runtime special-token post-processor updates are different tokenizer behaviors." }, { "left": "issue:43937", "right": "issue:44496", "accept": false, - "reason": "Invalid GenerationConfig for GLM-5 is not the same as an unrecognized model/config missing model_type." + "reason": "Invalid generation config and unrecognized model/config missing model_type are different validation failures." }, { "left": "issue:44117", "right": "issue:46328", "accept": false, - "reason": "Tokenizer mapping fallback behavior is unrelated to the SequenceFeatureExtractor.pad performance issue." + "reason": "Tokenizer mapping fallback and SequenceFeatureExtractor.pad performance issue are unrelated." }, { "left": "issue:43066", "right": "issue:43122", "accept": false, - "reason": "Both are v5 tokenizer regressions, but one is decoder type selection while the other is output/tokenization behavior; not the same concrete bug." + "reason": "Both are tokenizer-related v5 regressions, but the evidence only supports a loose family resemblance; there is no clear shared root cause or single mergeable fix." + }, + { + "left": "issue:44964", + "right": "issue:45278", + "accept": false, + "reason": "A specific multimodal model-load failure and a broad import-error report are not the same bug." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen3.5 loading failures with different model details; not enough to treat as one duplicate." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, + "evaluator_result": { + "accept": false, + "feedback": "The overall cluster summary is grounded: the packet does look heterogeneous, and using issue:45278 as a broad representative is reasonable. However, the soft-edge verdict for issue:43066 vs issue:43122 is too strong for the evidence in the packet. Both are tokenizer-related v5 regressions, but the titles only support a loose family resemblance, not a clearly shared root cause or a single mergeable bugfix. For this review, that positive accept is not conservative enough." + }, + "evaluator_used": true, + "retried": true, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -14681,12 +14602,12 @@ "issue:43986", "issue:44038", "issue:44075", + "issue:44117", "issue:44188", "issue:44220", "issue:44261", "issue:44263", "issue:44360", - "issue:44410", "issue:44483", "issue:44484", "issue:44485", @@ -14695,10 +14616,10 @@ "issue:44913", "issue:44933", "issue:44945", - "issue:44964", "issue:44977", + "issue:44987", "issue:45071", - "issue:45278", + "issue:45440", "issue:45464", "issue:45478", "issue:45542", @@ -14708,14 +14629,12 @@ "issue:46328" ], "soft_pairs": [ - "issue:44964|issue:45278", - "issue:44410|issue:45478", "issue:42222|issue:45542", "issue:42371|issue:44261", "issue:44075|issue:45656", "issue:44188|issue:46328", - "issue:44483|issue:45464", "issue:38617|issue:44220", + "issue:44483|issue:45464", "issue:43012|issue:43577", "issue:42175|issue:42222", "issue:44913|issue:46328", @@ -14739,228 +14658,230 @@ "issue:43874|issue:44263", "issue:43986|issue:44484", "issue:43673|issue:46328", - "issue:43976|issue:44945" + "issue:43976|issue:44945", + "issue:44117|issue:44987", + "issue:45440|issue:46172" ], - "prepared_review_unit_hash": "8a0f57038c9d7d1c7c55426fe8737a47a016472f99931a14b27c02a2abe5627f", + "prepared_review_unit_hash": "44bd56b631acc2f6e405709596e61c060d530ea61d5c92ce323f93cd336250b5", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 32, - "serialized_chars": 36932, - "estimated_input_tokens": 9233, - "estimated_eval_tokens": 18722 + "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": "The cluster is mostly a grab bag of unrelated Transformers issues with only superficial textual overlap (flash-attention, Qwen, import errors, dtype, etc.). I did not find any soft pair that clearly describes the same underlying bug/change, so all soft edges should be rejected.", - "confidence": 0.88, - "canonical_issue_reason": "issue:45464 is the most concrete and fully scoped bug report in the set, with a specific endpoint/model and a detailed reproduction, making it the best single anchor if one must be chosen.", + "summary": "This cluster is a grab-bag of Transformers bug reports spanning unrelated areas (model-specific Flash Attention issues, import failures, backend/install problems, config regressions, and optimizer/runtime bugs). None of the soft-similarity pairs look like true duplicates of the same underlying bug or change.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43097 is the best anchor for the set: it describes a concrete, user-visible v5 API break with downstream impact and is more representative of the version-regression theme than the narrower model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45464 is the strongest representative issue here because it is specific, user-facing, and more actionable than the broader or more ambiguous reports around it.", + "best_issue_reason": "issue:43097 is the strongest overall representative because it is a clear breaking-change report with broader applicability and higher triage value than the other, mostly isolated model/backend bugs.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44964", - "right": "issue:45278", - "accept": false, - "reason": "Different problems: one is a model-specific loading failure, the other is a broad post-upgrade import-error report." - }, - { - "left": "issue:44410", - "right": "issue:45478", - "accept": false, - "reason": "Both are Qwen-family issues, but they describe different model variants and different failure modes; not the same bug." - }, { "left": "issue:42222", "right": "issue:45542", "accept": false, - "reason": "Unrelated: vitpose model breakage vs TensorFlow backend/install packaging issue." + "reason": "Both involve TF/backend-related breakage, but one is a VitPose model failure and the other is a missing TensorFlow backend install path; different underlying bugs." }, { "left": "issue:42371", "right": "issue:44261", "accept": false, - "reason": "TF32 behavior settings and MLA layernorm precision config are distinct topics with no concrete shared code-path." + "reason": "TF32 settings guidance and an MLA RMS epsilon precision bug are unrelated precision topics, not the same defect." }, { "left": "issue:44075", "right": "issue:45656", "accept": false, - "reason": "Both involve optimizers, but one is about SGD args being ignored and the other about an extra optimizer step under deepspeed." + "reason": "One is about SGD args being ignored, the other about optimizer.step() being called twice under DeepSpeed; different optimizer-path failures." }, { "left": "issue:44188", "right": "issue:46328", "accept": false, - "reason": "Different subsystems entirely: torch.compile attention branching vs a SequenceFeatureExtractor.pad performance issue." + "reason": "Torch.compile attention-kernel divergence and SequenceFeatureExtractor.pad performance are unrelated code paths." }, { - "left": "issue:44483", - "right": "issue:45464", + "left": "issue:38617", + "right": "issue:44220", "accept": false, - "reason": "Similar API surface, but one is a general chat-completions rejection and the other is a specific streaming-inference failure on Qwen3.5-0.8B; not enough to treat as the same bug." + "reason": "An import error for configuration_utils and an audio feature extraction issue are different failures with no shared bug." }, { - "left": "issue:38617", - "right": "issue:44220", + "left": "issue:44483", + "right": "issue:45464", "accept": false, - "reason": "ImportError for a missing symbol and an audio feature-extraction issue are unrelated." + "reason": "Both affect chat/completions, but one is a v5.3 API rejection and the other is a Qwen3.5 streaming inference failure; not the same bug." }, { "left": "issue:43012", "right": "issue:43577", "accept": false, - "reason": "Both are dtype/precision related, but one is a compile-time warning and the other is a model-loading dtype mismatch in BLIP2." + "reason": "A PyTorch warning when compiling bf16 models is not the same as Blip2 dtype propagation staying float32." }, { "left": "issue:42175", "right": "issue:42222", "accept": false, - "reason": "Packaging/backend dependency issue vs a specific model breakage; not the same underlying defect." + "reason": "Both mention TensorFlow, but one is packaging/backend installation and the other is model breakage in VitPose." }, { "left": "issue:44913", "right": "issue:46328", "accept": false, - "reason": "GPTNeoX config reload bug and SequenceFeatureExtractor.pad performance regression are unrelated." + "reason": "GPTNeoX rotary_pct reload regression and a padding performance issue are unrelated." }, { "left": "issue:43097", "right": "issue:44493", "accept": false, - "reason": "Removed config API and unexpected position-id key errors are different regressions." + "reason": "Removed tie_embeddings_and_encoder_decoder and unexpected position-id keys are different config/state-dict regressions." }, { "left": "issue:44263", "right": "issue:44485", "accept": false, - "reason": "Both mention GLM/DSA-related code, but one is a torch.split indexing issue and the other is a RoPE implementation bug." + "reason": "A torch.split return-value issue and a GLM-5 RoPE implementation issue are different tensor logic bugs." }, { "left": "issue:43874", "right": "issue:44261", "accept": false, - "reason": "Different model/component families and different failure types; no clear shared bug." + "reason": "Missing get_number_of_image_patches and missing rms_norm_eps affect different multimodal/model components." }, { "left": "issue:41628", "right": "issue:45071", "accept": false, - "reason": "AutoImageProcessor import failure and PreTrainedConfig type-checking breakage are unrelated import/API regressions." + "reason": "AutoImageProcessor import failure and PreTrainedConfig type-checking regression are different API breakages." }, { "left": "issue:45071", "right": "issue:45542", "accept": false, - "reason": "Type-checking regression and undefined tf backend error are unrelated." + "reason": "Type-checking regression in PretrainedConfig is unrelated to the TensorFlow backend/install issue." }, { "left": "issue:27957", "right": "issue:43352", "accept": false, - "reason": "Both reference Flash Attention 2, but they affect different models with different support/status issues." + "reason": "XLMRoberta Flash Attention 2 support and NemotronHForCausalLM Flash Attention 2 support are different model-specific capability gaps." }, { "left": "issue:44263", "right": "issue:46172", "accept": false, - "reason": "Different model implementations and different tensor-shape/dtype problems." + "reason": "Both touch tensor transforms, but the DSA indexer split bug and the deepseek-v4 transpose bug are different code paths." }, { "left": "issue:44484", "right": "issue:45478", "accept": false, - "reason": "These are not the same bug: one asks about shard size defaults, the other is a Qwen3.5-Moe from_pretrained error." + "reason": "A shard-size question and a Qwen3.5-Moe from_pretrained error are unrelated." }, { "left": "issue:44360", "right": "issue:46172", "accept": false, - "reason": "DSA ReLU omission vs deepseek matmul/transposition bug; no shared concrete code-path." + "reason": "Missing ReLU in the DSA indexer and deepseek-v4 tensor transpose are distinct implementation bugs." }, { "left": "issue:43441", "right": "issue:44829", "accept": false, - "reason": "Both mention attention backends, but one is a model failing with FlashAttention in v5 RC and the other is degenerate training with flash_attention_3; different symptoms and likely different fixes." + "reason": "Ministral FlashAttention failure and flash_attention_3 degenerate training are both attention-related but not the same defect." }, { "left": "issue:42673", "right": "issue:44977", "accept": false, - "reason": "VRAM leak in multithreaded dataloader use vs incorrect generation with flash-attention are different problems." + "reason": "A VRAM leak in multithreaded loading is different from Qwen3.5 generation corruption under flash-attention." }, { "left": "issue:42673", "right": "issue:43906", "accept": false, - "reason": "No clear overlap beyond superficial reproducibility phrasing; the issues concern different models and behaviors." + "reason": "The VRAM leak and the isolated reproduction of issue 38071 are unrelated issues." }, { "left": "issue:42371", "right": "issue:42831", "accept": false, - "reason": "TF32 settings are unrelated to FineGrainedFP8 accuracy behavior." + "reason": "TF32 behavior control and FineGrainedFP8 accuracy regression are separate precision/accuracy concerns." }, { "left": "issue:43931", "right": "issue:44038", "accept": false, - "reason": "Both are Qwen3-VL-related, but one is a weight-shape mismatch on a specific model and the other is a broader v5/Qwen3-VL-Moe bug; not the same concrete issue." + "reason": "Qwen3-VL weight-shape mismatch is a loading mismatch, while Qwen3-VL-Moe on transformers 5.0 is a version-compatibility bug." }, { "left": "issue:43335", "right": "issue:45071", "accept": false, - "reason": "SwitchTransformers sparse-layer config bug and PreTrainedConfig type-checking regression are unrelated." + "reason": "A SwitchTransformers config edge case and a PretrainedConfig type-checking regression are different failures." }, { "left": "issue:43097", "right": "issue:43452", "accept": false, - "reason": "Removed tie_embeddings API and gguf_file loading breakage are different code paths." + "reason": "Removal of tie_embeddings_and_encoder_decoder and gguf_file breaking tokenizer/model loading are unrelated loading/API issues." }, { "left": "issue:44933", "right": "issue:46231", "accept": false, - "reason": "Import regression from image_utils vs documentation mismatch for set_training are unrelated." + "reason": "A nonexistent image_utils import and a documentation typo about set_training parameters are not the same bug." }, { "left": "issue:43874", "right": "issue:44263", "accept": false, - "reason": "Different model components and different missing-method vs tensor-split behavior." + "reason": "GLM46V image processor method absence and torch.split return values are unrelated component bugs." }, { "left": "issue:43986", "right": "issue:44484", "accept": false, - "reason": "Video-model/torchvision crash and save_pretrained shard-size question are unrelated." + "reason": "AutoProcessor crashing without torchvision and max_shard_size default behavior are unrelated." }, { "left": "issue:43673", "right": "issue:46328", "accept": false, - "reason": "Generation cache regression and feature-extractor padding performance issue are unrelated." + "reason": "Generation cache missing during chunked prefill and SequenceFeatureExtractor.pad performance are different areas." }, { "left": "issue:43976", "right": "issue:44945", "accept": false, - "reason": "Python-version compatibility problem and incorrect outputs under pipeline parallelism are unrelated." + "reason": "Python version support for Transformers 5.1.0 and incorrect output under pipeline parallelism are unrelated regressions." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and loading physical-intelligence/fast failing are different loader issues." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Native DeepseekV3MoE divergence from remote implementation and the deepseek-v4 comb.to/transpose bug are different DeepSeek-model issues, not the same code-path problem." } ] }, @@ -14997,6 +14918,7 @@ "issue:43208", "issue:43502", "issue:43550", + "issue:43825", "issue:43873", "issue:43906", "issue:43931", @@ -15008,6 +14930,7 @@ "issue:44261", "issue:44315", "issue:44351", + "issue:44368", "issue:44410", "issue:44484", "issue:44485", @@ -15015,27 +14938,23 @@ "issue:44704", "issue:44936", "issue:44964", - "issue:44987", "issue:45103", "issue:45200", "issue:45290", "issue:45341", "issue:45372", "issue:45375", - "issue:45440", "issue:45561", "issue:45588", + "issue:45712", "issue:45753", "issue:45859", "issue:46050", - "issue:46172", "issue:46298", "issue:46365", "issue:46369" ], "soft_pairs": [ - "issue:44117|issue:44987", - "issue:45440|issue:46172", "issue:43064|issue:44315", "issue:38617|issue:42757", "issue:43097|issue:45588", @@ -15044,8 +14963,8 @@ "issue:44410|issue:45375", "issue:42890|issue:46298", "issue:44038|issue:44410", - "issue:42673|issue:45200", "issue:43931|issue:44410", + "issue:42673|issue:45200", "issue:38617|issue:44351", "issue:44062|issue:44117", "issue:44261|issue:44485", @@ -15066,216 +14985,227 @@ "issue:43208|issue:43550", "issue:43066|issue:43906", "issue:43012|issue:43550", - "issue:42907|issue:44655" + "issue:42907|issue:44655", + "issue:43012|issue:44368", + "issue:45712|issue:46298", + "issue:43825|issue:44655", + "issue:42907|issue:45753", + "issue:43097|issue:46298" ], - "prepared_review_unit_hash": "8495a264941c35cdb37ec113ee7585aab689b3e5d6c1893d02bd68bd187bdf40", + "prepared_review_unit_hash": "3fec7ee8ad0f8d948c4a70e3e62a5931a933c2da6515c794a33b0edeca33ea75", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 33, - "serialized_chars": 37307, - "estimated_input_tokens": 9327, - "estimated_eval_tokens": 18910 + "soft_pair_count": 36, + "serialized_chars": 38370, + "estimated_input_tokens": 9593, + "estimated_eval_tokens": 19442 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The cluster is a loose collection of unrelated transformer issues, mostly model-specific bugs and import/load regressions. None of the soft-similarity pairs appear to describe the same underlying bug or change closely enough to merge.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "These items are largely unrelated regressions and edge-case bugs spread across different models, loading paths, and utility code. None of the soft pairs look like the same concrete bug/change, so I would not merge any of them as duplicates. Issue 43097 is the best cluster anchor because it is the broadest v5 regression report and has the strongest cross-cutting signal.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43097 is the most umbrella-like report in the set: a Transformers v5 breaking-change regression with multiple inbound references and discussion, making it the best anchor for the cluster.", "canonical_pr_reason": null, - "best_issue_reason": "No single issue is a clear canonical representative because the cluster is heterogeneous; if forced to inspect one first, issue 43873 is the most broadly scoped/open bug with multiple references, but it is not a duplicate of the others.", + "best_issue_reason": "issue:43097 is the most representative and broadly impactful issue here; it captures a cross-cutting v5 regression rather than a model-specific or one-off failure.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44117", - "right": "issue:44987", - "accept": false, - "reason": "Both involve loading/tokenizer mapping behavior, but one is a None-mapping assumption and the other is a model load failure for a specific repo; not the same bug." - }, - { - "left": "issue:45440", - "right": "issue:46172", - "accept": false, - "reason": "Different code paths and symptoms: remote DeepSeekV3 divergence vs a dtype/transpose bug in deepseek-v4." - }, { "left": "issue:43064", "right": "issue:44315", "accept": false, - "reason": "Training optimizer-state corruption with FSDP2/PEFT is unrelated to Liger Kernel not being applied via model_init." + "reason": "Different failures: optimizer-state/rank mismatch in FSDP2+PEFT vs Liger Kernel not applied with model_init." }, { "left": "issue:38617", "right": "issue:42757", "accept": false, - "reason": "Both are import errors, but they reference different missing symbols from different modules and likely different version mismatches." + "reason": "Both are import errors, but for different missing symbols in different modules; not the same underlying bug." }, { "left": "issue:43097", "right": "issue:45588", "accept": false, - "reason": "A removed config flag is unrelated to an AttributeError in flash_attention on sink-less models." + "reason": "Unrelated code paths: removed v5 embedding API vs flash_attention crash on sink-less models." }, { "left": "issue:43097", "right": "issue:46050", "accept": false, - "reason": "Different failures: config API removal vs broken quantized save/load after sinq serialization." + "reason": "Different regression classes: removed v5 API surface vs quantized model save/load corruption." }, { "left": "issue:43502", "right": "issue:43992", "accept": false, - "reason": "One is an unwanted network call despite local_files_only, the other is a missing encoder weight during model load." + "reason": "Both involve model loading, but one is unintended network access and the other is a missing state-dict weight." }, { "left": "issue:44410", "right": "issue:45375", "accept": false, - "reason": "Both hit Qwen3-family models, but one is missing attention projections and the other is a config field being dropped." + "reason": "Both are Qwen config/model issues, but they concern different missing fields and different model variants." }, { "left": "issue:42890", "right": "issue:46298", "accept": false, - "reason": "Test flakiness from missing set_seed is unrelated to a noisy logging message." + "reason": "Test flakiness from missing seeding is unrelated to a noisy logging message." }, { "left": "issue:44038", "right": "issue:44410", "accept": false, - "reason": "Same broad model family, but the concrete failures differ: Qwen3-VL-Moe loading versus qwen3next layer projection omission." + "reason": "Different Qwen problems: a v5 compatibility bug for Qwen3-VL-Moe vs missing projections in qwen3next layer 0." }, { - "left": "issue:42673", - "right": "issue:45200", + "left": "issue:43931", + "right": "issue:44410", "accept": false, - "reason": "VRAM leakage in multithreaded dataloading is unrelated to Gemma 4 requiring mm_token_type_ids defaults." + "reason": "Different model-loading failures: Qwen3-VL shape mismatch vs missing attention projections in a different model." }, { - "left": "issue:43931", - "right": "issue:44410", + "left": "issue:42673", + "right": "issue:45200", "accept": false, - "reason": "Different model-loading problems with different failure modes and likely different model configs." + "reason": "VRAM leak in multithreaded dataloaders is unrelated to Gemma 4's mm_token_type_ids defaulting behavior." }, { "left": "issue:38617", "right": "issue:44351", "accept": false, - "reason": "Both are missing-symbol import errors, but the symbols and affected dependencies differ." + "reason": "Both are import errors, but they involve different missing exports and different stack paths." }, { "left": "issue:44062", "right": "issue:44117", "accept": false, - "reason": "Tokenizer AddedToken argument duplication is a distinct bug from TOKENIZER_MAPPING_NAMES returning None." + "reason": "AddedToken argument collision is a tokenizer construction bug; TOKENIZER_MAPPING_NAMES returning None is a separate mapping/lookup issue." }, { "left": "issue:44261", "right": "issue:44485", "accept": false, - "reason": "A missing RMS norm epsilon in MLA is unrelated to GLM-5 RoPE implementation behavior." + "reason": "Different model math issues: missing rms_norm_eps precision drift vs GLM-5 RoPE implementation." }, { "left": "issue:45341", "right": "issue:45859", "accept": false, - "reason": "A testing_utils bug and a CPU-offload incompatibility in Qwen3_5 rotary embedding are unrelated." + "reason": "A testing_utils bug is unrelated to a CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding." }, { "left": "issue:43097", "right": "issue:43873", "accept": false, - "reason": "A removed embedding-tie API and a quantization offloading bug are not the same underlying problem." + "reason": "One is a removed v5 embedding API; the other is an offloading/quantization behavior bug." }, { "left": "issue:44964", "right": "issue:45341", "accept": false, - "reason": "Phi-4 multimodal loading failure is unrelated to a small testing_utils bug." + "reason": "Model loading for Phi-4 multimodal is unrelated to a testing_utils bug." }, { "left": "issue:44704", "right": "issue:45103", "accept": false, - "reason": "AutoProcessor kwarg propagation and auto_docstring annotation handling are separate subsystems and failures." + "reason": "Different layers entirely: AutoProcessor cached_file kwargs vs auto_docstring annotation parsing." }, { "left": "issue:43986", "right": "issue:44117", "accept": false, - "reason": "Missing torchvision handling in AutoProcessor is unrelated to tokenizer mapping returning None." + "reason": "Video AutoProcessor crash without torchvision is unrelated to tokenizer mapping returning None." }, { "left": "issue:45753", "right": "issue:46369", "accept": false, - "reason": "A Qwen3_5 infinite-generation loop and a SAM3 memory overflow are different model/runtime bugs." + "reason": "Different symptoms and subsystems: infinite generation loop vs memory overflow in a different model." }, { "left": "issue:42617", "right": "issue:45341", "accept": false, - "reason": "Running 3d_parallel.py and a testing utility bug do not share the same code path." + "reason": "Cannot run 3d_parallel.py is unrelated to a testing_utils bug." }, { "left": "issue:43986", "right": "issue:45372", "accept": false, - "reason": "Both are loader/processor related, but one depends on torchvision availability while the other is a mistral_common import regression." + "reason": "Missing torchvision during video processor loading is unrelated to a mistral_common ReasoningEffort import failure." }, { "left": "issue:43986", "right": "issue:45290", "accept": false, - "reason": "Processor loading without torchvision and chat-template tool-call handling are separate bugs." + "reason": "Processor loading without torchvision and chat-template tokenization on tool-call messages are different failures." }, { "left": "issue:43992", "right": "issue:44484", "accept": false, - "reason": "A missing embedding weight and the default max_shard_size value are unrelated." + "reason": "Missing embed_tokens.weight during from_pretrained is unrelated to the default max_shard_size question." }, { "left": "issue:44936", "right": "issue:46365", "accept": false, - "reason": "trainer.evaluate failing after train is distinct from an import conflict caused by datasets." + "reason": "Trainer.evaluate after train is a runtime-behavior bug; Trainer import failing due to datasets is an environment/import issue." }, { "left": "issue:43208", "right": "issue:43550", "accept": false, - "reason": "xLSTM training blockers and a Bamba torch.compile SDPA failure are unrelated model/backend issues." + "reason": "Different training/compile failures in different architectures; not the same code-path problem." }, { "left": "issue:43066", "right": "issue:43906", "accept": false, - "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue are not the same concrete bug." + "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue are not the same concrete defect." }, { "left": "issue:43012", "right": "issue:43550", "accept": false, - "reason": "A bfloat16 compile warning is unrelated to a model-specific torch.compile SDPA failure." + "reason": "A warning about bfloat16 compilation precision is not the same as a torch.compile SDPA failure." }, { "left": "issue:42907", "right": "issue:44655", "accept": false, - "reason": "Saving dequantized Ministrals/Devstrals and saving Pipeline objects are different save_pretrained problems." + "reason": "Saving dequantized models and saving Pipeline objects both involve save_pretrained, but they fail for different object types and paths." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Incorrect translation-task messaging in pipeline() is unrelated to pipeline serialization failure." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Left is dummy class cleanup/leakage, right is a logging-message complaint; no shared bug." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed v5 embedding API and noisy logging are unrelated." } ] }, @@ -15299,35 +15229,35 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ + "issue:36010", "issue:36683", "issue:38617", "issue:42490", "issue:42757", "issue:42898", - "issue:42907", "issue:42994", "issue:43012", "issue:43064", "issue:43066", - "issue:43097", "issue:43208", "issue:43299", "issue:43352", "issue:43475", "issue:43502", "issue:43526", + "issue:43611", "issue:43618", "issue:43761", "issue:43792", "issue:43824", - "issue:43825", "issue:43867", + "issue:43901", "issue:43937", "issue:44038", + "issue:44077", "issue:44162", "issue:44222", "issue:44230", - "issue:44368", "issue:44380", "issue:44483", "issue:44514", @@ -15337,23 +15267,18 @@ "issue:45276", "issue:45290", "issue:45478", + "issue:45584", "issue:45600", "issue:45698", - "issue:45712", - "issue:45753", "issue:46077", "issue:46123", "issue:46172", "issue:46178", "issue:46298", + "issue:46365", "issue:46421" ], "soft_pairs": [ - "issue:43012|issue:44368", - "issue:45712|issue:46298", - "issue:43825|issue:44655", - "issue:42907|issue:45753", - "issue:43097|issue:46298", "issue:44162|issue:46172", "issue:44938|issue:46178", "issue:43792|issue:46123", @@ -15376,211 +15301,219 @@ "issue:43618|issue:43761", "issue:42898|issue:43066", "issue:45276|issue:46077", - "issue:43352|issue:43937" + "issue:43352|issue:43937", + "issue:36010|issue:42757", + "issue:43901|issue:44655", + "issue:43611|issue:44655", + "issue:45584|issue:46365", + "issue:44077|issue:46365" ], - "prepared_review_unit_hash": "a0fbbcafeb1324224a10f4434290b1ff1c4ae4e3fe62671f634ae5285ec03a97", + "prepared_review_unit_hash": "8b967c324909c1823cb900c6cdb8e23d46f782c84c7c31dd1ec235b21923588f", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, "soft_pair_count": 28, - "serialized_chars": 35709, - "estimated_input_tokens": 8928, - "estimated_eval_tokens": 18112 + "serialized_chars": 35643, + "estimated_input_tokens": 8911, + "estimated_eval_tokens": 18078 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is highly heterogeneous: mostly unrelated Transformers v5 regressions, import errors, and model-specific bugs. The soft-similarity pairs do not look like true duplicates of the same underlying issue or change.", - "confidence": 0.88, - "canonical_issue_reason": "issue:45698 is the broadest load/save regression in the set and has the highest discussion activity, so it is the least-bad anchor for this mixed cluster.", + "summary": "This packet is still mostly a heterogeneous set of Transformers issues rather than one large duplicate cluster. The main representative issue is a broad open save/load regression (#45698), while the only soft edge that looks like the same underlying bug remains the pair of CLIP v5 output-regression issues.", + "confidence": 0.84, + "canonical_issue_reason": "No strong duplicate cluster emerges overall, so #45698 is a reasonable canonical issue because it is an open, user-facing save/load regression affecting custom module restoration. It is a useful representative anchor for triage, though it is not the most discussed issue in the packet.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45698 is open, user-facing, and centered on a general from_pretrained/save_pretrained codepath with meaningful discussion, making it the strongest representative issue here.", + "best_issue_reason": "#45698 is the best single issue anchor: it is open, broadly impactful, and describes a substantive save/load regression. #36683 has higher discussion activity, but it is a narrower config-specific failure rather than the better representative issue.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:43012", - "right": "issue:44368", - "accept": false, - "reason": "Different bugs: a PyTorch bfloat16 compilation warning vs a tie_word_embeddings warning during LoRA fine-tuning." - }, - { - "left": "issue:45712", - "right": "issue:46298", - "accept": false, - "reason": "Unrelated cleanup/CI issue about dummy classes vs a vague noisy logging complaint." - }, - { - "left": "issue:43825", - "right": "issue:44655", - "accept": false, - "reason": "Different code paths: pipeline error messaging vs save_pretrained for pipeline objects." - }, - { - "left": "issue:42907", - "right": "issue:45753", - "accept": false, - "reason": "Different symptoms and subsystems: saving dequantized models vs an infinite loop on a specific image in Qwen3_5." - }, - { - "left": "issue:43097", - "right": "issue:46298", - "accept": false, - "reason": "Removed API/documentation warning vs generic noisy logging; not the same underlying bug or change." - }, { "left": "issue:44162", "right": "issue:46172", "accept": false, - "reason": "ESM2 model breakage is unrelated to a deepseek-v4 tensor transpose issue." + "reason": "Different failures: ESM2 training breakage vs a DeepSeek-V4 tensor transpose/casting bug." }, { "left": "issue:44938", "right": "issue:46178", "accept": false, - "reason": "Python 3.14 import/load failure is unrelated to a backend mapping error for tensorflow_text." + "reason": "Unrelated problems: Python 3.14 load failure vs a backend-mapping validation error for tensorflow_text." }, { "left": "issue:43792", "right": "issue:46123", "accept": false, - "reason": "Whisper runtime failure and MaskGenerationPipeline batch-dropping are distinct issues." + "reason": "Whisper inference failure and MaskGenerationPipeline batch-dropping are separate code paths and symptoms." }, { "left": "issue:44222", "right": "issue:45478", "accept": false, - "reason": "Both involve MoE/FP8 model loading or saving, but they concern different failure modes and likely different code paths." + "reason": "Both involve Qwen3.5/MoE FP8, but one is save_pretrained and the other is from_pretrained; not clearly the same concrete bug." }, { "left": "issue:42490", "right": "issue:43012", "accept": false, - "reason": "Local model path slash warning is unrelated to a bfloat16 precision warning." + "reason": "A path-parsing import warning and a bf16 precision warning are unrelated regressions." }, { "left": "issue:44483", "right": "issue:44514", "accept": false, - "reason": "OpenAI chat-completions API rejection is unrelated to a Qwen2.5-VL processor batching crash." + "reason": "Both touch chat APIs, but one is request acceptance at /v1/chat/completions and the other is a specific chat-template crash on batched input." }, { "left": "issue:43012", "right": "issue:44683", "accept": false, - "reason": "Different problems: precision warning during compile vs compiled flex_attention failure on newer torch." + "reason": "Float32 precision warning during bf16 compilation is not the same as a compiled flex_attention failure on newer torch." }, { "left": "issue:43208", "right": "issue:43526", "accept": false, - "reason": "xLSTM training bugs and BeitImageProcessorFast reduce_labels behavior do not match." + "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated model-component issues." }, { "left": "issue:38617", "right": "issue:43502", "accept": false, - "reason": "ImportError from a removed symbol is unrelated to local_files_only still triggering network requests." + "reason": "An import error for layer_type_validation and offline-mode network requests are separate bugs." }, { "left": "issue:42757", "right": "issue:43824", "accept": false, - "reason": "Different import failures: is_offline_mode removal vs missing Qwen2_5_VL model export." + "reason": "Both are import errors, but they concern different missing symbols and likely different package/export breakages." }, { "left": "issue:45600", "right": "issue:46298", "accept": false, - "reason": "Auto-mapping references to removed Sam3LiteText configs is not the same as a noisy logging message." + "reason": "Broken auto_mappings references and a noisy logging message are not the same underlying issue." }, { "left": "issue:44380", "right": "issue:46421", "accept": false, - "reason": "GPT2 attention scaling/backend behavior is unrelated to cache lazy initialization recompilation." + "reason": "GPT2 attention-scaling behavior and cache lazy initialization recompilation are different backend behaviors." }, { "left": "issue:45698", "right": "issue:46172", "accept": false, - "reason": "Wrong custom module loading after save_pretrained is unrelated to a deepseek-v4 dtype/transpose bug." + "reason": "Custom module restoration after save_pretrained is unrelated to DeepSeek-V4 tensor conversion/transposition." }, { "left": "issue:44038", "right": "issue:44230", "accept": false, - "reason": "Related Qwen3/VL/MoE theme, but one is a generic load bug and the other is FP8 inference support; not the same concrete fix." + "reason": "Both mention Qwen3-VL/MoE, but one is a general Transformers 5 regression and the other is fp8 inference support." }, { "left": "issue:44483", "right": "issue:45290", "accept": false, - "reason": "API request validation and assistant tool-call tokenization are different failure modes." + "reason": "Both involve chat handling, but the reported failures are distinct and not clearly the same concrete bug." }, { "left": "issue:43299", "right": "issue:43867", "accept": false, - "reason": "Qwen3VL MoE loading breakage and sorted state_dict loading failure are different loading bugs." + "reason": "A base_model_prefix loading regression and a state_dict-sorting load error are different model-loading problems." }, { "left": "issue:43064", "right": "issue:44380", "accept": false, - "reason": "Trainer/FSDP2 optimizer-state corruption is unrelated to GPT2 attention scaling under SDPA/FlashAttention." + "reason": "Distributed training optimizer-state corruption and GPT2 attention scaling are unrelated." }, { "left": "issue:36683", "right": "issue:43475", "accept": false, - "reason": "Gemma3Config missing vocab_size and SAM3 vision encoder missing fpn_position_embeddings are different model attribute regressions." + "reason": "Missing vocab_size on Gemma3Config and missing fpn_position_embeddings on Sam3VisionEncoderOutput are different attribute regressions." }, { "left": "issue:42994", "right": "issue:44655", "accept": false, - "reason": "Quantized model saving failure and pipeline save_pretrained failure are different save paths and not clearly one bug." + "reason": "Quantized model save failure and pipeline save_pretrained failure are distinct save paths with different objects involved." }, { "left": "issue:43618", "right": "issue:43761", - "accept": false, - "reason": "Both are CLIP v5 regressions, but one is missing attentions assignment and the other missing hidden_states; adjacent but distinct." + "accept": true, + "reason": "Both look like the same CLIP v5 regression in output handling: one drops attentions and the other drops hidden_states when outputs are requested." }, { "left": "issue:42898", "right": "issue:43066", "accept": false, - "reason": "Tokenizer cleanup behavior change and wrong tokenizer decoder type are separate tokenizer issues." + "reason": "Tokenizer cleanup behavior and wrong decoder type are separate v5 tokenizer issues." }, { "left": "issue:45276", "right": "issue:46077", "accept": false, - "reason": "Gemma4 embedding resize bug is unrelated to causal-mask construction for bidirectional attention." + "reason": "Gemma4 embedding resize behavior and bidirectional-attention masking are separate model-behavior bugs." }, { "left": "issue:43352", "right": "issue:43937", "accept": false, - "reason": "Nemotron Flash Attention 2 support and GLM-5 GenerationConfig validation are unrelated." + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig validation are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Different missing imports from different modules; same general symptom but not the same bug." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "Documentation mismatch for TextClassificationPipeline is unrelated to pipeline serialization failure." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Loading models with base_model_prefix and saving Pipeline objects are different regressions." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation and Trainer import interference from datasets are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "An optional post_init contract issue in patchtsmixer is not the same as the Trainer import problem." } ] }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, + "evaluator_result": { + "accept": false, + "feedback": "The overall clustering is mostly conservative, but the canonical issue rationale is not grounded: #45698 does not have the highest discussion activity in the packet (#36683 has 42 vs. #45698's 15). Please revise that claim and keep the representative-issue wording more cautious (e.g., a substantive open save/load regression), rather than asserting it as the most discussed issue." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, @@ -15597,11 +15530,12 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:36010", + "issue:28282", + "issue:33453", + "issue:36296", "issue:36683", "issue:41669", "issue:42175", - "issue:42757", "issue:42890", "issue:43064", "issue:43299", @@ -15609,15 +15543,13 @@ "issue:43441", "issue:43502", "issue:43525", - "issue:43611", + "issue:43575", "issue:43618", "issue:43867", "issue:43881", - "issue:43901", "issue:43976", "issue:43992", "issue:44075", - "issue:44077", "issue:44117", "issue:44242", "issue:44246", @@ -15628,8 +15560,10 @@ "issue:44625", "issue:44655", "issue:44855", + "issue:44857", "issue:44871", "issue:44938", + "issue:44977", "issue:45216", "issue:45310", "issue:45357", @@ -15637,20 +15571,15 @@ "issue:45478", "issue:45479", "issue:45561", - "issue:45584", "issue:45753", "issue:46077", "issue:46082", "issue:46132", "issue:46202", + "issue:46285", "issue:46365" ], "soft_pairs": [ - "issue:36010|issue:42757", - "issue:43901|issue:44655", - "issue:43611|issue:44655", - "issue:45584|issue:46365", - "issue:44077|issue:46365", "issue:44871|issue:46365", "issue:43867|issue:45216", "issue:43502|issue:44485", @@ -15679,240 +15608,259 @@ "issue:43976|issue:44938", "issue:43064|issue:44492", "issue:43867|issue:45357", - "issue:43992|issue:45478" + "issue:43992|issue:45478", + "issue:36296|issue:43575", + "issue:43064|issue:44655", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:44655|issue:44938", + "issue:43618|issue:44857", + "issue:43618|issue:44977" ], - "prepared_review_unit_hash": "5f61adceaa89f8e84032b8215c5813cb7ef5651f75605efeb303dd0331e845be", + "prepared_review_unit_hash": "fb204f5856bd736eac935102c756e73b2141232227aa7383554bcbacbf9a292c", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 47, - "item_count": 47, - "soft_pair_count": 34, - "serialized_chars": 37041, - "estimated_input_tokens": 9261, - "estimated_eval_tokens": 18778 + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38140, + "estimated_input_tokens": 9535, + "estimated_eval_tokens": 19326 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly noisy: the items are a mix of unrelated Transformers bugs, regressions, and docs/perf reports. I found one strong duplicate pair around the Qwen3.5 MoE `from_pretrained` regression; the rest look like distinct issues.", - "confidence": 0.84, - "canonical_issue_reason": "issue:45310 is the clearest canonical issue because it has the exact same failure mode and scope as issue:45478 (same title, same Qwen3.5 MoE `from_pretrained` regression), and it is the earlier report.", + "summary": "This cluster is mostly a grab bag of Transformers issues around loading/saving, config mismatches, attention backends, import/performance, and packaging. I don\u2019t see any soft pair that clearly reflects the same underlying bug/change, so I\u2019d keep them separate.", + "confidence": 0.82, + "canonical_issue_reason": "issue:43867 is the broadest user-facing model-loading failure in the set and the closest thing to a reusable umbrella anchor, though it is not a true duplicate of the others.", "canonical_pr_reason": null, - "best_issue_reason": "issue:45310 is the best issue anchor for the cluster\u2019s only true duplicate pair: it is specific, versioned, and directly matches the later duplicate report.", + "best_issue_reason": "issue:43867 is the most generic and representative loading-regression report here, with a concrete reproduction and enough breadth to serve as the cluster\u2019s best issue anchor.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:36010", - "right": "issue:42757", - "accept": false, - "reason": "Both are import errors, but for different symbols/modules (`GenerationMixin` vs `is_offline_mode`); no sign of the same underlying bug." - }, - { - "left": "issue:43901", - "right": "issue:44655", - "accept": false, - "reason": "One is a docs/behavior mismatch for `return_all_scores`; the other is about `Pipeline.save_pretrained`. Different features and failure modes." - }, - { - "left": "issue:43611", - "right": "issue:44655", - "accept": false, - "reason": "`base_model_prefix` loading regression and pipeline saving are unrelated code paths." - }, - { - "left": "issue:45584", - "right": "issue:46365", - "accept": false, - "reason": "Whisper empty-transcription generation vs `Trainer` import conflict with `datasets`; different bugs entirely." - }, - { - "left": "issue:44077", - "right": "issue:46365", - "accept": false, - "reason": "`patchtsmixer` post-init handling and `Trainer` import resolution are not the same defect." - }, { "left": "issue:44871", "right": "issue:46365", "accept": false, - "reason": "Gemma eos-token config inconsistency is unrelated to the `Trainer` import issue." + "reason": "Gemma eos-token inconsistency vs Trainer import conflict with datasets; unrelated bugs." }, { "left": "issue:43867", "right": "issue:45216", "accept": false, - "reason": "Both involve loading/saving, but they target different models and different failures (`state_dict` ordering vs Qwen3.5 checkpoint regression)." + "reason": "Generic load error when state_dict is sorted vs Qwen3.5 save_pretrained checkpoint regression; different failure modes." }, { "left": "issue:43502", "right": "issue:44485", "accept": false, - "reason": "Local-files-only network leakage and GLM-5 RoPE behavior are unrelated." + "reason": "local_files_only network leak vs GLM-5 RoPE discussion; not the same bug." }, { "left": "issue:44625", "right": "issue:46077", "accept": false, - "reason": "`num_labels` propagation in Qwen3.5 and Gemma4 attention-mask construction are different issues." + "reason": "Qwen3.5 num_labels propagation vs Gemma4 bidirectional attention-mask bug; different code paths." }, { "left": "issue:46077", "right": "issue:46082", "accept": false, - "reason": "Both are config-related, but one is about Gemma4 bidirectional attention masks and the other about Llama head-dim validation; not the same bug." + "reason": "Gemma4 attention mask issue vs LlamaConfig head_dim validation; unrelated." }, { "left": "issue:36683", "right": "issue:43525", "accept": false, - "reason": "Missing `vocab_size` on Gemma3Config vs missing `pad_token_id` on Llama4Config are separate model-specific config regressions." + "reason": "Missing vocab_size vs missing pad_token_id; similar shape, different model-specific config defects." }, { "left": "issue:43618", "right": "issue:45216", "accept": false, - "reason": "CLIP attentions assignment and Qwen3.5 save_pretrained corruption are different regressions." + "reason": "CLIP attentions API/structure change vs Qwen3.5 save_pretrained regression; different subsystems." }, { "left": "issue:42175", "right": "issue:46365", "accept": false, - "reason": "Backend dependency packaging and `Trainer` import conflicts are unrelated." + "reason": "Package/backend install problem vs Trainer import issue tied to datasets; not the same underlying bug." }, { "left": "issue:44380", "right": "issue:46132", "accept": false, - "reason": "Attention scaling being ignored in SDPA/FlashAttention is not the same as `AttentionInterface.register` changing function behavior." + "reason": "GPT2 scaling ignored in SDPA/FlashAttention vs AttentionInterface.register behavior change; distinct attention bugs." }, { "left": "issue:36683", "right": "issue:44242", "accept": false, - "reason": "Gemma3 config attribute missing and MoE load-balancing loss handling are different problems." + "reason": "Gemma3 config attribute missing vs load-balancing loss omission; unrelated." }, { "left": "issue:36683", "right": "issue:43881", "accept": false, - "reason": "Both are loading/config failures, but for different models and symptoms; not a duplicate." + "reason": "Gemma3 config attribute missing vs glm-4v-9b loading failure; different models and causes." }, { "left": "issue:43299", "right": "issue:45753", "accept": false, - "reason": "Qwen3VL MoE model-loading breakage and Qwen3_5 infinite-loop-on-image are unrelated." + "reason": "Qwen3VL MoE loading break vs Qwen3_5 infinite-loop generation on one image; different symptoms." }, { "left": "issue:43352", "right": "issue:43976", "accept": false, - "reason": "FlashAttention support for one model and Python version compatibility for the whole package are different issues." + "reason": "FlashAttention unsupported for Nemotron vs Python-version compatibility issue; unrelated." }, { "left": "issue:43992", "right": "issue:45310", "accept": false, - "reason": "UMT5 missing `embed_tokens.weight` and Qwen3.5 MoE `from_pretrained` failure are different load paths and model families." + "reason": "UMT5Encoder missing embed_tokens.weight vs Qwen3.5 MoE from_pretrained error; different models and failure mechanisms." }, { "left": "issue:44075", "right": "issue:46365", "accept": false, - "reason": "SGD argument handling and `Trainer` import conflict are unrelated." + "reason": "SGD args ignored vs Trainer import problem; unrelated." }, { "left": "issue:43352", "right": "issue:45478", "accept": false, - "reason": "Unsupported FlashAttention on Nemotron and Qwen3.5 MoE load failure are different bugs." + "reason": "Same broad FlashAttention area, but different model/repro and not clearly one fixable code-path bug." }, { "left": "issue:44355", "right": "issue:44855", "accept": false, - "reason": "Compiled-file import errors vs a Python 3.13 `torch.jit.script` parsing/IndentationError are different root causes." + "reason": "Compiled-file errors vs DebertaV2 import parse failure on Python 3.13; different root causes." }, { "left": "issue:43441", "right": "issue:44380", "accept": false, - "reason": "Both mention attention backends, but one is a model-specific FlashAttention failure and the other is an attention-scaling config regression." + "reason": "Ministral FlashAttention failure vs GPT2 attention scaling bug; distinct model/backend issues." }, { "left": "issue:43867", "right": "issue:45375", "accept": false, - "reason": "A generic load-model ordering error and a missing config field dropped by `@strict` are not the same defect." + "reason": "Generic load error vs Qwen3_5 vision config field being dropped; not the same defect." }, { "left": "issue:43867", "right": "issue:46202", "accept": false, - "reason": "State-dict ordering failure and Gemma pipeline roundtrip failure have different triggers and code paths." + "reason": "Model load failure vs Gemma pipeline roundtrip problem on \u2581; unrelated." }, { "left": "issue:36683", "right": "issue:44117", "accept": false, - "reason": "Missing config attribute on Gemma3 and tokenizer-mapping `None` handling are unrelated." + "reason": "Missing config attribute vs tokenizer-mapping returning None; different bug classes." }, { "left": "issue:43352", "right": "issue:44380", "accept": false, - "reason": "Unsupported FlashAttention for one model and ignored GPT-2 scaling config are different issues." + "reason": "FlashAttention support/error for one model vs GPT2 scaling being ignored; not the same issue." }, { "left": "issue:44380", "right": "issue:45479", "accept": false, - "reason": "Attention scaling in backends and zero-loss classification behavior are unrelated." + "reason": "Attention scaling bug vs degenerate zero-loss classification; unrelated." }, { "left": "issue:41669", "right": "issue:44246", "accept": false, - "reason": "Both concern import slowness, but 41669 is a specific `import *` cleanup proposal while 44246 is a generic slowdown report; not enough evidence of the same underlying bug." + "reason": "Import-* cleanup/import-time regression vs intermittent import slowness; plausible theme overlap, but not the same concrete bug/change." }, { "left": "issue:42890", "right": "issue:45561", "accept": false, - "reason": "One is a flaky integration test needing `set_seed`; the other is a pytest-xdist race in patched testing utils." + "reason": "Test seeding flakiness vs xdist race on captured_info.txt; separate test infra issues." }, { "left": "issue:43976", "right": "issue:44938", "accept": false, - "reason": "Python 3.9+/3.10+ compatibility and Python 3.14 import failure are different version-specific issues." + "reason": "Python version support mismatch vs Python 3.14 import failure; different compatibility problems." }, { "left": "issue:43064", "right": "issue:44492", "accept": false, - "reason": "FSDP2/PEFT optimizer-state corruption and a cache-strategy typo are unrelated." + "reason": "FSDP2/PEFT optimizer-state bug vs cache-strategy typo; unrelated." }, { "left": "issue:43867", "right": "issue:45357", "accept": false, - "reason": "State-dict sorting failure and incorrect Qwen3.5 visual encoder keys in save_pretrained are both save/load themed but distinct bugs." + "reason": "Model load error vs incorrect visual-encoder keys saved in Qwen3.5; different checkpoint problems." }, { "left": "issue:43992", "right": "issue:45478", "accept": false, - "reason": "Different models and different missing artifacts; not the same `from_pretrained` bug." + "reason": "UMT5Encoder weight-missing issue vs Qwen3.5 MoE from_pretrained error; different model families and repros." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor parallel training bug vs tp loading OOM; one is a correctness bug, the other a memory-capacity issue." + }, + { + "left": "issue:43064", + "right": "issue:44655", + "accept": false, + "reason": "Optimizer-state mismatch during training vs pipeline save_pretrained failure; unrelated." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch dependency ImportError vs tokenizer-loading regression; not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "CLIPOutput attentions removal vs CLIP text-model structure inconsistency; related family, but distinct changes." + }, + { + "left": "issue:44655", + "right": "issue:44938", + "accept": false, + "reason": "Pipeline save_pretrained bug vs Python 3.14 import failure; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "CLIPOutput attentions API issue vs LwDetrImageLoss AMP/CUDA crash; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44977", + "accept": false, + "reason": "CLIP attentions regression vs Qwen3.5 flash-attention generation loop; unrelated." } ] }, @@ -15937,26 +15885,26 @@ "cluster_id": "cluster-10000-459", "nodes": [ "issue:27957", - "issue:28282", "issue:30333", "issue:30990", "issue:31795", "issue:33453", "issue:34689", "issue:36296", + "issue:38617", "issue:42548", + "issue:42757", "issue:42831", "issue:42890", "issue:43010", - "issue:43064", "issue:43299", "issue:43502", - "issue:43575", "issue:43576", - "issue:43618", "issue:43723", + "issue:43824", "issue:43827", "issue:43874", + "issue:43976", "issue:44062", "issue:44112", "issue:44222", @@ -15969,30 +15917,23 @@ "issue:44556", "issue:44655", "issue:44749", - "issue:44857", + "issue:44908", "issue:44938", "issue:44961", "issue:44964", - "issue:44977", "issue:45071", "issue:45125", "issue:45278", "issue:45375", "issue:45440", + "issue:45468", "issue:45478", "issue:45600", + "issue:45656", "issue:45850", - "issue:46132", - "issue:46285" + "issue:46132" ], "soft_pairs": [ - "issue:36296|issue:43575", - "issue:43064|issue:44655", - "issue:28282|issue:33453", - "issue:43618|issue:46285", - "issue:44655|issue:44938", - "issue:43618|issue:44857", - "issue:43618|issue:44977", "issue:44360|issue:45440", "issue:43502|issue:44062", "issue:42890|issue:44964", @@ -16021,240 +15962,273 @@ "issue:33453|issue:34689", "issue:44315|issue:44360", "issue:43299|issue:44556", - "issue:27957|issue:33453" + "issue:27957|issue:33453", + "issue:42757|issue:43576", + "issue:43576|issue:45071", + "issue:43010|issue:45656", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:43010|issue:43976", + "issue:43576|issue:44938", + "issue:44908|issue:44961" ], - "prepared_review_unit_hash": "e0a9109f4d5a62f8e5d7a5917fbc2951ba9ff685dda87f67269573a188b89c83", + "prepared_review_unit_hash": "3e260e6649eb61047cda00e84edb951ee4761afa2723d04a80ef6818cde5a7bc", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 36, - "serialized_chars": 38137, - "estimated_input_tokens": 9535, - "estimated_eval_tokens": 19326 + "soft_pair_count": 38, + "serialized_chars": 38630, + "estimated_input_tokens": 9658, + "estimated_eval_tokens": 19572 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "The candidate set is heterogeneous: most items are separate v5/model-loading regressions, import/export breakages, CI/test flakes, or unrelated training/runtime bugs. The soft edges are generally only broad topical similarities, not the same underlying bug/change, so none are safe duplicates/merge candidates.", - "confidence": 0.81, - "canonical_issue_reason": "If forced to pick one issue as the closest umbrella, #45278 is the best representative of the v5 upgrade/import-breakage family. However, the overall set is too mixed for a single true canonical issue across all items.", + "summary": "These items are mostly unrelated issue reports; the only recurring pattern is broad v5 regressions/import/load failures, but none of the soft pairs look like the same underlying bug.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45278 is the clearest umbrella-style report here: it clearly states a broad version-regression symptom ('many import errors after update from 4.57.0 to 5.5.0') and is more representative than the narrowly scoped one-off bugs.", "canonical_pr_reason": null, - "best_issue_reason": "#45278 is the strongest global representative because it is broad, version-specific, and covers several of the other v5 regression-style reports better than the more narrowly scoped issues. It is still only a partial fit, not a perfect canonical.", + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it is specific enough to be understandable yet broad enough to represent the shared v5 breakage theme seen in nearby reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:36296", - "right": "issue:43575", - "accept": false, - "reason": "Both mention tensor parallelism, but one is a training bug and the other is a load-time OOM for a specific model; not the same code-path." - }, - { - "left": "issue:43064", - "right": "issue:44655", - "accept": false, - "reason": "FSDP/PEFT optimizer-state corruption and pipeline save_pretrained are unrelated behaviors." - }, - { - "left": "issue:28282", - "right": "issue:33453", - "accept": false, - "reason": "Different failures: missing PyTorch import vs tokenizer-loading regression." - }, - { - "left": "issue:43618", - "right": "issue:46285", - "accept": false, - "reason": "Both are CLIP-related regressions, but one is about attentions on outputs and the other about text_model structure; different bugs." - }, - { - "left": "issue:44655", - "right": "issue:44938", - "accept": false, - "reason": "Pipeline save_pretrained and Python 3.14 import/load failure are unrelated." - }, - { - "left": "issue:43618", - "right": "issue:44857", - "accept": false, - "reason": "CLIP attentions regression and LwDetrImageLoss AMP crash are different model/path issues." - }, - { - "left": "issue:43618", - "right": "issue:44977", - "accept": false, - "reason": "CLIP output field regression and Qwen3.5 flash-attention generation bug are unrelated." - }, { "left": "issue:44360", "right": "issue:45440", "accept": false, - "reason": "Different model/component failures; one is a DSA indexer ReLU discussion, the other is DeepSeekV3MoE implementation divergence." + "reason": "Completely different problems: a DSA indexer/ReLU discussion vs a DeepSeekV3MoE implementation divergence." }, { "left": "issue:43502", "right": "issue:44062", "accept": false, - "reason": "local_files_only network leakage and AddedToken keyword duplication are unrelated." + "reason": "One is about unwanted network access with local_files_only=True; the other is a tokenizers.AddedToken argument error." }, { "left": "issue:42890", "right": "issue:44964", "accept": false, - "reason": "A flaky SAM HQ integration test and a Phi-4 multimodal model-loading failure are unrelated." + "reason": "Seed-related flaky tests for SamHQ are unrelated to Phi-4 multimodal model loading failures." }, { "left": "issue:42890", "right": "issue:44112", "accept": false, - "reason": "Both are CI/test issues, but they concern different tests and different root causes." + "reason": "A missing set_seed() in a test is unrelated to a GraniteSpeech CI device-override failure." }, { "left": "issue:42890", "right": "issue:45600", "accept": false, - "reason": "Test flakiness due to missing set_seed and a broken auto_mappings reference are not the same bug." + "reason": "Test flakiness in SamHQ does not match an auto_mappings.py CI break from removed configs." }, { "left": "issue:44222", "right": "issue:45125", "accept": false, - "reason": "Both involve MoE/tensor-parallel areas, but one is FP8 save_pretrained and the other is missing _tp_plan; different problems." + "reason": "Both mention MoE/FP8-adjacent areas, but one is about save_pretrained with FP8 and the other about missing _tp_plan for tensor parallelism." }, { "left": "issue:43576", "right": "issue:46132", "accept": false, - "reason": "Broken env CLI command and AttentionInterface.register behavior change are unrelated." + "reason": "A broken CLI env command and AttentionInterface.register behavior changes are different code paths and symptoms." }, { "left": "issue:45125", "right": "issue:45478", "accept": false, - "reason": "Both touch Qwen3.5 MoE loading, but one is a missing tensor-parallel plan and the other is a broader from_pretrained failure; not enough evidence of the same fix." + "reason": "Same model family, but one is missing tensor-parallel metadata while the other is a from_pretrained regression; not the same concrete bug." }, { "left": "issue:42548", "right": "issue:44351", "accept": false, - "reason": "Both are import errors, but for different exported symbols and likely different causes." + "reason": "Both are import errors, but they involve different symbols and likely different breakages, so they are not the same bug." }, { "left": "issue:44355", "right": "issue:44749", "accept": false, - "reason": "Compiled Python file errors and a data-filtering slowdown after upgrade are unrelated." + "reason": "Running compiled Python files errors and slowed filtering after upgrade are unrelated." }, { "left": "issue:30333", "right": "issue:33453", "accept": false, - "reason": "MLFlow reporting on failed training jobs and tokenizer-loading regression are unrelated." + "reason": "MLflow reporting on failed training jobs is unrelated to tokenizer-loading regression." }, { "left": "issue:44387", "right": "issue:44556", "accept": false, - "reason": "Int4 quantization memory regression and checkpoint reload failure are different bugs." + "reason": "Memory growth under int4 quantization is a different failure mode from checkpoint reload incompatibility." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is missing tensor-parallel plan data while the other is a missing vision config field silently dropped by strict parsing." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "A no_grad decoration request for cache updates is unrelated to import errors after upgrading to v5.5.0." }, { "left": "issue:43723", "right": "issue:44556", "accept": false, - "reason": "Tokenizer loading in v5 and checkpoint reload after upgrade are both loading issues, but not the same concrete failure." + "reason": "Tokenizer loading issues and checkpoint reload failures are both load-related but not the same underlying problem." }, { "left": "issue:30990", "right": "issue:31795", "accept": false, - "reason": "Sentence Transformers hanging on load and confusing forward docs are unrelated." + "reason": "Sentence Transformers loading hangs and documentation confusion about model.forward arguments are unrelated." }, { "left": "issue:43502", "right": "issue:44222", "accept": false, - "reason": "Network requests despite local_files_only and FP8 MoE save_pretrained are unrelated." + "reason": "Local-files-only network leakage is unrelated to FP8 save_pretrained behavior." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No meaningful overlap: offline-loading network calls vs an unrelated placeholder issue." }, { "left": "issue:42831", "right": "issue:43874", "accept": false, - "reason": "FP8 accuracy and missing image-processor method are unrelated." + "reason": "An accuracy issue in FineGrainedFP8 is not the same as a missing image-patch helper causing an AttributeError." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "A cache update no_grad change is unrelated to a vague general complaint that v5 breaks too many models." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow reporting and tensor-parallel training bugs are different subsystems and failure modes." }, { "left": "issue:44355", "right": "issue:45071", "accept": false, - "reason": "Compiled-file errors and PretrainedConfig type-checking breakage are unrelated." + "reason": "Compiled-file execution errors are unrelated to PretrainedConfig type-checking regressions." + }, + { + "left": "issue:43010", + "right": "issue:44368", + "accept": false, + "reason": "Cache update semantics have nothing to do with a tie_word_embeddings warning in Qwen3.5 LoRA fine-tuning." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline() and an unrelated placeholder issue are not duplicates." }, { "left": "issue:44355", "right": "issue:44655", "accept": false, - "reason": "Compiled Python file failures and pipeline save_pretrained are unrelated." + "reason": "Compiled Python file errors and inability to save Pipeline objects are different issues." }, { "left": "issue:33453", "right": "issue:34689", "accept": false, - "reason": "Tokenizer-loading regression and Llama 3.2 Vision model-loading breakage are different concrete failures." + "reason": "Tokenizer-loading regression and Llama 3.2 model-loading breakage are both loading failures, but they are not the same concrete bug." }, { "left": "issue:44315", "right": "issue:44360", "accept": false, - "reason": "Liger Kernel not applied with model_init and DSA indexer lacking ReLU are unrelated." + "reason": "Liger Kernel application with model_init is unrelated to a DSA indexer missing ReLU." }, { "left": "issue:43299", "right": "issue:44556", "accept": false, - "reason": "Qwen3VL MoE loading breakage and checkpoint reload failure are both version-regression reports, but not the same bug." + "reason": "Qwen3VL MoE loading under dev0 and checkpoint reload after upgrading to v5.2/v5.3 are distinct regression reports." }, { "left": "issue:27957", "right": "issue:33453", "accept": false, - "reason": "Flash Attention 2 support for XLM-Roberta and tokenizer-loading regression are unrelated." + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to tokenizer loading regression." }, { - "left": "issue:44355", + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "An import error from huggingface_hub is not the same as the transformers env CLI being broken." + }, + { + "left": "issue:43576", "right": "issue:45071", "accept": false, - "reason": "Duplicate entry omitted in source data? No separate soft pair beyond the one above; if considered, still reject for unrelated causes." + "reason": "Broken env CLI behavior and PretrainedConfig type-checking are unrelated." }, { - "left": "issue:45125", - "right": "issue:45375", + "left": "issue:43010", + "right": "issue:45656", + "accept": false, + "reason": "Cache no_grad decoration and double optimizer steps under deepspeed are different training-path bugs." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "The env command failure is unrelated to Qwen2.5-VL import failures." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and an ImportError for layer_type_validation are unrelated." + }, + { + "left": "issue:44961", + "right": "issue:45468", "accept": false, - "reason": "Missing _tp_plan and missing deepstack_visual_indexes are both Qwen3.5 MoE-related, but they concern different schema/code-path issues." + "reason": "An unrelated placeholder issue does not match the Gemma-4 audio positional encoding bug." }, { "left": "issue:43010", - "right": "issue:45278", + "right": "issue:43976", "accept": false, - "reason": "Cache/layer update no_grad decoration and broad import errors after upgrade are unrelated." + "reason": "Cache update no_grad behavior is unrelated to Python 3.9/3.10 support problems." }, { - "left": "issue:43723", - "right": "issue:44556", + "left": "issue:43576", + "right": "issue:44938", "accept": false, - "reason": "Tokenizer.from_pretrained failure and checkpoint reload failure are different loading regressions." + "reason": "The transformers env command and Python 3.14 load failure are different issues." }, { - "left": "issue:43502", - "right": "issue:44222", + "left": "issue:44908", + "right": "issue:44961", "accept": false, - "reason": "Network requests with local_files_only and FP8 MoE save_pretrained are unrelated." + "reason": "Scheduler kwargs handling has no relation to the unrelated placeholder issue." } ] }, @@ -16278,15 +16252,11 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:31795", "issue:36296", - "issue:38617", "issue:41628", "issue:41669", "issue:42548", - "issue:42757", "issue:42890", - "issue:43010", "issue:43352", "issue:43408", "issue:43525", @@ -16294,7 +16264,7 @@ "issue:43572", "issue:43576", "issue:43673", - "issue:43824", + "issue:43874", "issue:43901", "issue:43976", "issue:44038", @@ -16305,38 +16275,33 @@ "issue:44295", "issue:44297", "issue:44315", + "issue:44351", "issue:44373", "issue:44393", "issue:44556", "issue:44623", + "issue:44704", "issue:44908", - "issue:44938", "issue:44961", "issue:45003", "issue:45071", "issue:45092", "issue:45125", "issue:45230", + "issue:45278", + "issue:45341", "issue:45362", - "issue:45468", + "issue:45478", "issue:45542", - "issue:45656", "issue:45753", "issue:45850", "issue:46095", - "issue:46209", - "issue:46218" - ], - "soft_pairs": [ - "issue:42757|issue:43576", - "issue:43576|issue:45071", - "issue:43010|issue:45656", - "issue:43576|issue:43824", - "issue:31795|issue:38617", - "issue:44961|issue:45468", - "issue:43010|issue:43976", - "issue:43576|issue:44938", - "issue:44908|issue:44961", + "issue:46143", + "issue:46144", + "issue:46209", + "issue:46218" + ], + "soft_pairs": [ "issue:44222|issue:44230", "issue:36296|issue:43901", "issue:41628|issue:42548", @@ -16365,264 +16330,266 @@ "issue:44295|issue:45230", "issue:43576|issue:44908", "issue:43976|issue:44393", - "issue:44908|issue:45071" + "issue:44908|issue:45071", + "issue:44704|issue:44908", + "issue:41669|issue:44351", + "issue:45278|issue:45850", + "issue:41669|issue:45341", + "issue:44393|issue:45478", + "issue:43874|issue:45850", + "issue:46143|issue:46144", + "issue:43976|issue:46218" ], - "prepared_review_unit_hash": "e07372facb502ea7f4ca9f957d2e08923974152e1daaa8b38e4b65487291f572", + "prepared_review_unit_hash": "83a279587a3c88f19e45bf5f3a45a09d3cd416b5bc11d4d9649e7b6cb52b17d9", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { "node_count": 48, "item_count": 48, - "soft_pair_count": 38, - "serialized_chars": 38388, - "estimated_input_tokens": 9597, - "estimated_eval_tokens": 19450 + "soft_pair_count": 37, + "serialized_chars": 38060, + "estimated_input_tokens": 9515, + "estimated_eval_tokens": 19286 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is heterogeneous: most items are unrelated import/export, lazy-loading, version-upgrade, or model-specific regressions. The soft pairs share broad themes at best, but not the same concrete bug or change, so none should be merged as duplicates.", - "confidence": 0.93, - "canonical_issue_reason": "issue:44273 is the broadest, most central issue here ('Lazy loading is not working properly') and best matches the recurring import/lazy-loading regressions seen in a few nearby items.", + "summary": "Mostly a noisy mix of unrelated Transformers v5 regression reports, with a few broad import/lazy-loading complaints and many model-specific bugs. The soft pairs share only superficial subsystem overlap, not the same underlying defect or change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45278 is the best umbrella issue here: it explicitly captures the broad \"many import errors after update from 4.57.0 to 5.5.0\" regression theme that most closely matches the cluster\u2019s dominant pattern.", "canonical_pr_reason": null, - "best_issue_reason": "issue:44273 is the best umbrella representative for the cluster, but only loosely; the set still spans many distinct bugs rather than one underlying failure mode.", + "best_issue_reason": "issue:45278 is the most representative of the cluster because it generalizes the repeated import/load breakages without being as vague as the catch-all complaints.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:42757", - "right": "issue:43576", - "accept": false, - "reason": "Different failures: a missing huggingface_hub symbol versus a broken transformers env command." - }, - { - "left": "issue:43576", - "right": "issue:45071", - "accept": false, - "reason": "Unrelated problems: env command behavior versus PreTrainedConfig type-checking." - }, - { - "left": "issue:43010", - "right": "issue:45656", - "accept": false, - "reason": "Cache update no_grad semantics and deepspeed double optimizer step are different code paths." - }, - { - "left": "issue:43576", - "right": "issue:43824", - "accept": false, - "reason": "env command issue and Qwen2.5-VL import error are not the same bug." - }, - { - "left": "issue:31795", - "right": "issue:38617", - "accept": false, - "reason": "Doc confusion about forward args versus an import error for layer_type_validation." - }, - { - "left": "issue:44961", - "right": "issue:45468", - "accept": false, - "reason": "Placeholder/irrelevant issue versus a Gemma-4 audio positional-encoding bug." - }, - { - "left": "issue:43010", - "right": "issue:43976", - "accept": false, - "reason": "torch.no_grad cache updates and Python version compatibility are unrelated." - }, - { - "left": "issue:43576", - "right": "issue:44938", - "accept": false, - "reason": "env command breakage and Python 3.14 load failure are different symptoms and fixes." - }, - { - "left": "issue:44908", - "right": "issue:44961", - "accept": false, - "reason": "Scheduler kwargs bug versus unrelated placeholder issue." - }, { "left": "issue:44222", "right": "issue:44230", "accept": false, - "reason": "Both mention FP8/moe, but one is save_pretrained and the other is inference support; not the same concrete bug." + "reason": "Both involve fp8/moe, but one is about save_pretrained behavior and the other about inference support; different code paths and fixes." }, { "left": "issue:36296", "right": "issue:43901", "accept": false, - "reason": "Tensor parallel training bug versus pipeline documentation mismatch." + "reason": "Tensor parallel training bug vs a pipeline doc mismatch; no shared underlying bug." }, { "left": "issue:41628", "right": "issue:42548", "accept": false, - "reason": "Both are import errors, but for different missing exports and different root causes." + "reason": "Both are import errors, but for different exported symbols and likely different breakages." }, { "left": "issue:44315", "right": "issue:46209", "accept": false, - "reason": "Liger kernel/model_init application is unrelated to FP8 inference on B200." + "reason": "Liger Kernel application in model_init is unrelated to FP8 inference failure on B200." }, { "left": "issue:36296", "right": "issue:43526", "accept": false, - "reason": "Tensor parallel training bug versus BEiT reduce_labels returning the wrong shape." + "reason": "Tensor parallel training vs BeitImageProcessorFast label reduction; unrelated subsystems and symptoms." }, { "left": "issue:44623", "right": "issue:45230", "accept": false, - "reason": "Processor save_pretrained missing files is too vague to match the unrelated generic bug report." + "reason": "A processor save_pretrained file omission is not the same as a vague general bug report." }, { "left": "issue:44273", "right": "issue:45542", "accept": false, - "reason": "Lazy loading regression and missing TensorFlow backend handling are different issues." + "reason": "Lazy loading problems and tensorboard/TF backend detection are different bugs." }, { "left": "issue:44038", "right": "issue:45125", "accept": false, - "reason": "General Qwen3-VL/Moe problem versus a specific missing tensor-parallel plan." + "reason": "Qwen3-VL-Moe bug vs missing _tp_plan for tensor parallelism; related model family but different concrete failures." }, { "left": "issue:44393", "right": "issue:45071", "accept": false, - "reason": "Qwen3-VL 2D box output issue versus config type-checking breakage." + "reason": "Qwen3-VL 2D box output issue vs PretrainedConfig type checking regression; not the same defect." }, { "left": "issue:44273", "right": "issue:45003", "accept": false, - "reason": "Lazy loading malfunction and unsafe sys.modules access are related only loosely, not the same bug." + "reason": "Lazy loading behavior and sys.modules access safety are distinct implementation problems." }, { "left": "issue:44373", "right": "issue:44961", "accept": false, - "reason": "Wrong docstring for position_ids versus unrelated placeholder issue." + "reason": "Docstring wording for position_ids vs an unrelated one-word issue title; no evidence of shared bug." }, { "left": "issue:44230", "right": "issue:46095", "accept": false, - "reason": "FP8 inference support and a weight conversion-script question are not the same change." + "reason": "FP8 inference support for Qwen models is not the same as asking about a weight conversion script." }, { "left": "issue:44393", "right": "issue:45753", "accept": false, - "reason": "Both are Qwen3-VL/image related, but one is bbox hallucination/error and the other is an infinite loop on a specific image." + "reason": "Both involve Qwen3.5-VL behavior, but one is 2D box output/hallucination and the other is an infinite loop on a specific image; too different to merge." }, { "left": "issue:43576", "right": "issue:44556", "accept": false, - "reason": "env command failure and old checkpoint reload incompatibility are distinct version regressions." + "reason": "Broken `transformers env` command and checkpoint reload incompatibility after upgrade are separate issues." }, { "left": "issue:43525", "right": "issue:45850", "accept": false, - "reason": "Missing pad_token_id in one config is not the same as a broad 'v5 breaks many models' complaint." + "reason": "A missing pad_token_id on Llama4Config is a specific config bug, while the other issue is a broad complaint about many models breaking." }, { "left": "issue:41669", "right": "issue:43673", "accept": false, - "reason": "Import-star removal/performance work and missing GenerationMixin cache are different fixes." + "reason": "Import * usage slowing imports is not the same as missing GenerationMixin cache during chunked prefill." }, { "left": "issue:44556", "right": "issue:45092", "accept": false, - "reason": "Both concern v5 loading, but one is generic checkpoint reload and the other is remote-code/meta-init incompatibility; not one concrete bug." + "reason": "Both are upgrade/meta-init related, but one is checkpoint reload incompatibility and the other is remote-code checkpoint incompatibility; different concrete failures." }, { "left": "issue:43572", "right": "issue:45850", "accept": false, - "reason": "StableLmConfig pad token field regression is specific; the other is a vague umbrella complaint." + "reason": "Missing pad_token_idx in one config is a specific regression; the other is a broad multi-model complaint." }, { "left": "issue:45362", "right": "issue:45850", "accept": false, - "reason": "Qwen3.5 chat crash is a specific model bug, not the generic v5-breakage report." + "reason": "Qwen3.5-35B chat crash is model-specific, not the same as a generalized v5 breakage report." }, { "left": "issue:41669", "right": "issue:46218", "accept": false, - "reason": "Import-star cleanup/performance issue versus module import failure for Qwen3_5ForCausalLM." + "reason": "Import * usage from models and inability to import Qwen3_5ForCausalLM are both import-related, but not the same bug or fix." }, { "left": "issue:43408", "right": "issue:44315", "accept": false, - "reason": "sam3 tracker config warning and Liger Kernel application are unrelated." + "reason": "sam3_video/sam3_tracker warning is a model-type mismatch, unrelated to Liger Kernel application in model_init." }, { "left": "issue:42890", "right": "issue:44908", "accept": false, - "reason": "Missing set_seed in tests and inverse_sqrt scheduler kwargs are unrelated." + "reason": "Missing set_seed in a SAM HQ integration test is unrelated to inverse_sqrt scheduler kwargs handling." }, { "left": "issue:43352", "right": "issue:44393", "accept": false, - "reason": "Flash Attention support gating for Nemotron versus Qwen3-VL bbox output errors." + "reason": "Flash Attention 2 support error for NemotronH vs Qwen3-VL 2D box output failure are different model and failure modes." }, { "left": "issue:44246", "right": "issue:44273", "accept": false, - "reason": "Slow import sometimes and lazy loading broken are related only at a high level, not the same bug." + "reason": "Import sometimes taking long and lazy loading not working are related only at a very broad loading-performance level, not the same underlying defect." }, { "left": "issue:44297", "right": "issue:45230", "accept": false, - "reason": "Tokenizer save_pretrained metadata mismatch is unrelated to the generic bug report." + "reason": "Tokenizer class metadata mismatch on save_pretrained is distinct from a generic bug report." }, { "left": "issue:44295", "right": "issue:45230", "accept": false, - "reason": "position_ids buffer access error is unrelated to the generic bug report." + "reason": "Reading position_ids after buffer registration is a specific tensor/buffer issue, not the same as a generic bug report." }, { "left": "issue:43576", "right": "issue:44908", "accept": false, - "reason": "env command breakage and scheduler kwargs loss do not share a code path." + "reason": "Environment command breakage and scheduler kwargs ignored are unrelated features." }, { "left": "issue:43976", "right": "issue:44393", "accept": false, - "reason": "Python version support regression versus Qwen3-VL bounding-box inference failure." + "reason": "Python version compatibility for transformers 5.1.0 is unrelated to a Qwen3-VL output error." }, { "left": "issue:44908", "right": "issue:45071", "accept": false, - "reason": "Scheduler parameter handling and config type-checking are unrelated." + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are separate issues." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor.cached_file kwargs propagation and scheduler kwargs handling are different code paths." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "import * usage removal from models and missing HybridCache import are both import-related, but not the same defect." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "A concrete set of import errors after upgrade is not the same as a vague broad complaint that v5 breaks too many models." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "import * performance/import issue and a testing_utils bug are unrelated." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3-VL 2D box hallucination/error and Qwen3.5 Moe from_pretrained error are different model failures." + }, + { + "left": "issue:43874", + "right": "issue:45850", + "accept": false, + "reason": "Missing get_number_of_image_patches in one image processor is a specific API omission, not the same as a broad v5 complaint." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Not passing kwargs through RoFormer methods and attention implementation not using the attention interface are related to RoFormer, but they are different code-path problems." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python version support regression is unrelated to a module import failure for Qwen3_5ForCausalLM." } ] }, @@ -16646,10 +16613,13 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ + "issue:30333", "issue:31515", "issue:31795", "issue:33453", "issue:34689", + "issue:36296", + "issue:38617", "issue:41669", "issue:42371", "issue:42913", @@ -16658,52 +16628,40 @@ "issue:43352", "issue:43502", "issue:43519", + "issue:43525", + "issue:43572", "issue:43576", "issue:43606", "issue:43673", "issue:43704", "issue:43827", - "issue:43874", - "issue:43976", "issue:44077", + "issue:44230", "issue:44261", "issue:44263", "issue:44273", "issue:44315", "issue:44351", "issue:44360", - "issue:44393", "issue:44485", "issue:44492", "issue:44623", "issue:44655", - "issue:44704", "issue:44829", "issue:44861", "issue:44908", + "issue:44998", "issue:45230", - "issue:45278", + "issue:45290", "issue:45310", - "issue:45341", "issue:45468", - "issue:45478", "issue:45850", "issue:46018", "issue:46095", - "issue:46143", - "issue:46144", "issue:46172", "issue:46218" ], "soft_pairs": [ - "issue:44704|issue:44908", - "issue:41669|issue:44351", - "issue:45278|issue:45850", - "issue:41669|issue:45341", - "issue:44393|issue:45478", - "issue:43874|issue:45850", - "issue:46143|issue:46144", - "issue:43976|issue:46218", "issue:44077|issue:44861", "issue:43673|issue:44492", "issue:43827|issue:44908", @@ -16737,288 +16695,339 @@ "issue:44261|issue:45468", "issue:44351|issue:46218", "issue:43352|issue:46218", - "issue:42913|issue:43066" + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44861|issue:45290", + "issue:43704|issue:44998", + "issue:43525|issue:44861", + "issue:31515|issue:38617", + "issue:43572|issue:44861", + "issue:43066|issue:44998", + "issue:41669|issue:44998", + "issue:44998|issue:46218", + "issue:44230|issue:44623", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230" ], - "prepared_review_unit_hash": "bce46756fc02794200c65616d7be0be83ac4f8f3b0e7a8074e86e9a1395ca5da", + "prepared_review_unit_hash": "0a5267d7f3f7e4397c6f64edfa008ea6e990570eded90500ebee3e5bc0521f79", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 48, - "item_count": 48, - "soft_pair_count": 42, - "serialized_chars": 39503, - "estimated_input_tokens": 9876, - "estimated_eval_tokens": 20008 + "node_count": 47, + "item_count": 47, + "soft_pair_count": 49, + "serialized_chars": 41024, + "estimated_input_tokens": 10256, + "estimated_eval_tokens": 20768 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "This cluster is mostly a heterogeneous mix of unrelated Transformers regressions, docs issues, and model-specific bugs, with only a loose theme around v5 loading/import breakages. I don\u2019t see any soft pair that clearly describes the same underlying bug.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 45278 is the best anchor if one must be chosen: it is the broadest concrete report of v5 import/loading breakage. The rest of the set is too mixed to form a true duplicate cluster.", + "summary": "These items do not form a real duplicate cluster; the soft-similarity links mostly connect unrelated issues that only share broad vocabulary (v5, loading, cache, tokenizer, etc.). I\u2019d keep one concrete, broadly scoped issue as the representative anchor and reject all soft edges.", + "confidence": 0.74, + "canonical_issue_reason": "No true canonical duplicate set emerges here; issue 41669 is the strongest standalone anchor because it has a concrete, broadly impactful regression with clearer scope and more discussion than the vague or highly model-specific reports.", "canonical_pr_reason": null, - "best_issue_reason": "45278 is the most representative issue for the recurring \u2018v5 breaks imports/model loading\u2019 complaints, but it is still only a loose umbrella rather than a true duplicate target.", + "best_issue_reason": "Issue 41669 is the best single issue to retain as a representative reference point: it is specific, actionable, and has broader ecosystem impact than the other mostly narrow or ambiguous reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44704", - "right": "issue:44908", - "accept": false, - "reason": "AutoProcessor kwargs forwarding to cached_file and inverse_sqrt scheduler kwargs are unrelated code paths." - }, - { - "left": "issue:41669", - "right": "issue:44351", - "accept": false, - "reason": "Import-* cleanup/performance work is not the same bug as a HybridCache import failure." - }, - { - "left": "issue:45278", - "right": "issue:45850", - "accept": false, - "reason": "Broad v5 breakage complaint and vague 'too many models' report overlap only at a high level, not a single concrete defect." - }, - { - "left": "issue:41669", - "right": "issue:45341", - "accept": false, - "reason": "Model import* cleanup and a testing_utils typo are unrelated." - }, - { - "left": "issue:44393", - "right": "issue:45478", - "accept": false, - "reason": "Different models and different failures: Qwen3-VL 2D box output vs Qwen3.5 MoE from_pretrained error." - }, - { - "left": "issue:43874", - "right": "issue:45850", - "accept": false, - "reason": "Missing GLM46V image-patch API and vague v5 breakage are not the same bug." - }, - { - "left": "issue:46143", - "right": "issue:46144", - "accept": false, - "reason": "One is kwargs passthrough, the other is attention interface usage; related area but different defects." - }, - { - "left": "issue:43976", - "right": "issue:46218", - "accept": false, - "reason": "Python version compatibility and a Qwen3_5 import error are unrelated." - }, { "left": "issue:44077", "right": "issue:44861", "accept": false, - "reason": "Optional post_init policy and _tied_weights_keys AttributeError are different issues." + "reason": "Different bugs: post_init validation vs AttributeError in tied-weight key handling." }, { "left": "issue:43673", "right": "issue:44492", "accept": false, - "reason": "Chunked prefill cache missing and a cache-strategy typo do not describe the same bug." + "reason": "Chunked prefill cache behavior is unrelated to a typo in cache strategy docs." }, { "left": "issue:43827", "right": "issue:44908", "accept": false, - "reason": "Docs still referencing pipeline() and scheduler kwargs handling are unrelated." + "reason": "Documentation cleanup after pipeline removals is unrelated to scheduler kwargs handling." }, { "left": "issue:31795", "right": "issue:33453", "accept": false, - "reason": "Forward-doc confusion and tokenizer loading regression are different problems." + "reason": "Doc confusion about forward args is not the same as a tokenizer loading regression." }, { "left": "issue:41669", "right": "issue:44492", "accept": false, - "reason": "Import-* cleanup is unrelated to a typo in cache strategies." + "reason": "Import-star performance issue is unrelated to a cache-strategy typo." }, { "left": "issue:44315", "right": "issue:44829", "accept": false, - "reason": "Liger kernel not applied in model_init and flash_attention_3 training degeneration are different code-path bugs." + "reason": "Model-init Liger Kernel application and flash-attention training degeneration are different failures." }, { "left": "issue:43576", "right": "issue:44273", "accept": false, - "reason": "Broken env command and lazy loading failure are unrelated." + "reason": "Broken env command is unrelated to lazy loading behavior." }, { "left": "issue:44655", "right": "issue:44861", "accept": false, - "reason": "Pipeline save_pretrained and tied-weight key handling are different save/model internals." + "reason": "Pipeline save_pretrained support is unrelated to tied-weight key crashes." }, { "left": "issue:44360", "right": "issue:45468", "accept": false, - "reason": "DSA indexer ReLU omission and Gemma-4 audio positional encoding are unrelated model-specific defects." + "reason": "DSA indexer ReLU discussion is unrelated to Gemma audio positional encoding." }, { "left": "issue:45230", "right": "issue:45310", "accept": false, - "reason": "A generic bug report is not the same as the specific Qwen3.5 MoE from_pretrained error." + "reason": "Generic bug report placeholder vs a specific Qwen3.5 MoE from_pretrained error." }, { "left": "issue:42371", "right": "issue:44492", "accept": false, - "reason": "TF32 API settings and cache-strategy typo are unrelated." + "reason": "TF32 API migration is unrelated to a cache-strategy typo." }, { "left": "issue:44273", "right": "issue:44908", "accept": false, - "reason": "Lazy loading and inverse_sqrt scheduler kwargs are different subsystems." + "reason": "Lazy loading and inverse_sqrt scheduler kwargs are different code paths." }, { "left": "issue:43576", "right": "issue:46218", "accept": false, - "reason": "Env command failure and Qwen3_5 import error are unrelated." + "reason": "Env command breakage is unrelated to a Qwen3_5 import failure." }, { "left": "issue:43827", "right": "issue:44492", "accept": false, - "reason": "Pipeline docs cleanup and cache-strategy typo do not match." + "reason": "Pipeline-removal docs issue is unrelated to a cache-strategy typo." }, { "left": "issue:43519", "right": "issue:44485", "accept": false, - "reason": "Qwen3VL timestamp calculation and GLM-5 RoPE implementation are different model bugs." + "reason": "Timestamp calculation in Qwen3VL is unrelated to GLM-5 RoPE implementation." }, { "left": "issue:43519", "right": "issue:44908", "accept": false, - "reason": "Timestamp calculation in a processor and scheduler kwargs handling are unrelated." + "reason": "Timestamp logic and scheduler kwargs are unrelated bugs." }, { "left": "issue:46018", "right": "issue:46172", "accept": false, - "reason": "Both mention DeepSeek-V4, but shared-expert gating and comb.transpose dtype handling are distinct bugs." + "reason": "DeepSeek shared-expert gating and a tensor transpose issue are different model bugs." }, { "left": "issue:41669", "right": "issue:43704", "accept": false, - "reason": "Import-* cleanup and a VRAM leak in dataloader threads are unrelated." + "reason": "Import-star removal performance work is unrelated to a VRAM leak in dataloader threads." }, { "left": "issue:31795", "right": "issue:34689", "accept": false, - "reason": "Documentation confusion and Llama 3.2 model-loading breakage are not the same issue." + "reason": "Forward-doc confusion is unrelated to Llama 3.2 model loading regression." }, { "left": "issue:31515", "right": "issue:31795", "accept": false, - "reason": "Slow checkpoint loading and forward API documentation confusion are different problems." + "reason": "Slow checkpoint loading is not the same as forward-doc confusion." }, { "left": "issue:43704", "right": "issue:44485", "accept": false, - "reason": "VRAM leak and RoPE implementation bug are unrelated." + "reason": "VRAM leak in threaded loading is unrelated to GLM-5 RoPE behavior." }, { "left": "issue:44485", "right": "issue:45468", "accept": false, - "reason": "Different models and different positional-encoding bugs; not mergeable as one fix." + "reason": "Both are model-specific math issues, but they affect different architectures and code paths." }, { "left": "issue:45850", "right": "issue:46218", "accept": false, - "reason": "Vague v5 breakage complaints and a specific Qwen3_5 import failure are not the same defect." + "reason": "Broad v5 breakage report is not the same as a specific Qwen3_5 import error." }, { "left": "issue:42371", "right": "issue:43704", "accept": false, - "reason": "TF32 settings and VRAM leak are unrelated." + "reason": "TF32 settings API and VRAM leak reports are unrelated." }, { "left": "issue:43298", "right": "issue:46095", "accept": false, - "reason": "Continuous batching audio support and DeepSeek-V4 weight conversion are unrelated." + "reason": "Continuous batching support for audio models is unrelated to DeepSeek-V4 weight conversion." }, { "left": "issue:43502", "right": "issue:43519", "accept": false, - "reason": "local_files_only still causing requests and timestamp calculation are different loading/processor bugs." + "reason": "local_files_only network leakage is unrelated to Qwen3VL timestamp computation." }, { "left": "issue:43519", "right": "issue:44623", "accept": false, - "reason": "Timestamp calculation and processor.save_pretrained missing files are unrelated." + "reason": "Timestamp math and missing processor.save_pretrained files are different bugs." }, { "left": "issue:43519", "right": "issue:43704", "accept": false, - "reason": "Processor timestamp math and VRAM leak are unrelated." + "reason": "Timestamp calculation bug is unrelated to VRAM leakage in threaded loading." }, { "left": "issue:44263", "right": "issue:45468", "accept": false, - "reason": "torch.split return handling in GlmMoeDsaIndexer and Gemma-4 audio positional encoding are different defects." + "reason": "torch.split return-value concerns are unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "RMS norm epsilon precision mismatch is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "HybridCache import failure and Qwen3_5 import failure are different symbols and likely different fixes." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention 2.0 support gating is unrelated to an import error." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both mention tokenizer v5, but one is broad behavior drift and the other is a specific decoder-type issue." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug is unrelated to the placeholder 'Unemployment' issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight key handling and chat-template tool-call crash are different failures." + }, + { + "left": "issue:43704", + "right": "issue:44998", + "accept": false, + "reason": "VRAM leak issue is unrelated to the placeholder report." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is unrelated to tied-weight key crashes." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading is unrelated to an ImportError for layer_type_validation." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_idx after v5 update is unrelated to tied-weight key crashes." + }, + { + "left": "issue:43066", + "right": "issue:44998", + "accept": false, + "reason": "Tokenizer decoder type bug is unrelated to the placeholder report." + }, + { + "left": "issue:41669", + "right": "issue:44998", + "accept": false, + "reason": "Import-star removal/performance issue is unrelated to the placeholder report." + }, + { + "left": "issue:44998", + "right": "issue:46218", + "accept": false, + "reason": "Placeholder issue is unrelated to a specific Qwen3_5 import failure." }, { - "left": "issue:43606", - "right": "issue:45468", + "left": "issue:44230", + "right": "issue:44623", "accept": false, - "reason": "CPU offload device mismatch and Gemma-4 audio positional encoding are unrelated." + "reason": "FP8 support work is unrelated to missing files in processor.save_pretrained." }, { - "left": "issue:44261", - "right": "issue:45468", + "left": "issue:30333", + "right": "issue:31515", "accept": false, - "reason": "MLA q_a_layernorm epsilon precision and Gemma-4 audio positional encoding are different bugs." + "reason": "MLFlow reporting on failed training jobs is unrelated to slow checkpoint loading." }, { - "left": "issue:44351", - "right": "issue:46218", + "left": "issue:31515", + "right": "issue:36296", "accept": false, - "reason": "Both are import errors, but they involve different missing symbols/models and different fixes." + "reason": "Checkpoint loading slowdown is unrelated to tensor-parallel training bugs." }, { - "left": "issue:43352", - "right": "issue:46218", + "left": "issue:44230", + "right": "issue:45310", "accept": false, - "reason": "Flash Attention 2 support error and Qwen3_5 import failure are unrelated." + "reason": "FP8 inference support and Qwen3.5 MoE from_pretrained failure are different model features/issues." }, { - "left": "issue:42913", - "right": "issue:43066", + "left": "issue:43298", + "right": "issue:44230", "accept": false, - "reason": "Tokenizer behavior changes and wrong decoder type are both tokenizer-related, but not the same concrete bug." + "reason": "Continuous batching for audio models is unrelated to Qwen3-VL/Qwen3.5 FP8 support." } ] }, @@ -17042,51 +17051,17 @@ "model": "gpt-5.4-mini?service_tier=flex", "cluster_id": "cluster-10000-459", "nodes": [ - "issue:30333", - "issue:31515", - "issue:33453", - "issue:36296", "issue:37934", - "issue:38617", - "issue:41669", "issue:42936", - "issue:43066", "issue:43240", - "issue:43298", - "issue:43525", - "issue:43572", "issue:43668", - "issue:43704", "issue:43717", "issue:43813", "issue:44031", - "issue:44230", - "issue:44623", - "issue:44861", - "issue:44908", - "issue:44998", - "issue:45290", - "issue:45310", "issue:46018", - "issue:46144", - "issue:46218" + "issue:46144" ], "soft_pairs": [ - "issue:44908|issue:44998", - "issue:44861|issue:45290", - "issue:43704|issue:44998", - "issue:43525|issue:44861", - "issue:31515|issue:38617", - "issue:43572|issue:44861", - "issue:43066|issue:44998", - "issue:41669|issue:44998", - "issue:44998|issue:46218", - "issue:44230|issue:44623", - "issue:30333|issue:31515", - "issue:31515|issue:36296", - "issue:44230|issue:45310", - "issue:31515|issue:33453", - "issue:43298|issue:44230", "issue:43717|issue:46144", "issue:42936|issue:46018", "issue:43668|issue:44031", @@ -17100,196 +17075,106 @@ "issue:37934|issue:43668", "issue:42936|issue:43668" ], - "prepared_review_unit_hash": "e76e54d7cce4d42f431358277b1f5dee225c743823ac19b3526f3c873e5dae70", + "prepared_review_unit_hash": "7f8c2d8e9e743f6bd8b9b4dd7f41af4b3b975f92d4731fb03800125def60fd79", "component_budget": { "node_count": 459, "item_count": 459, "soft_pair_count": 1688, - "serialized_chars": 875291, - "estimated_input_tokens": 218823, - "estimated_eval_tokens": 437902 + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 }, "budget": { - "node_count": 28, - "item_count": 28, - "soft_pair_count": 27, - "serialized_chars": 23694, - "estimated_input_tokens": 5924, - "estimated_eval_tokens": 12104 + "node_count": 9, + "item_count": 9, + "soft_pair_count": 12, + "serialized_chars": 8610, + "estimated_input_tokens": 2153, + "estimated_eval_tokens": 4562 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These issues are a loose collection of unrelated bugs across tokenizers, configs, model implementations, training/runtime behavior, plus one clearly off-topic item. The soft-similarity links are too broad to treat as true duplicates, so no merges are safe here.", - "confidence": 0.95, - "canonical_issue_reason": null, + "summary": "The cluster is a loose mix of unrelated model, config, tokenizer, and utility issues. None of the soft pairs look like the same underlying bug or change, so no duplicates should be merged here.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44031 is the clearest standalone bug report: it describes a concrete, user-facing regression with a specific version trigger and broad tokenizer impact.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Issue 44031 is the strongest representative of the set because it is specific, actionable, and broadly impactful compared with the more model-local or narrow reports.", "best_pr_reason": null, "soft_edge_verdicts": [ - { - "left": "issue:44908", - "right": "issue:44998", - "accept": false, - "reason": "Different topics: lr_scheduler_kwargs/timescale vs an off-topic spam issue. No shared bug or code path." - }, - { - "left": "issue:44861", - "right": "issue:45290", - "accept": false, - "reason": "Unrelated failures: tied-weights helper crash vs chat template/tool-call crash." - }, - { - "left": "issue:43704", - "right": "issue:44998", - "accept": false, - "reason": "VRAM leak in Qwen3ForCausalLM is unrelated to the off-topic issue." - }, - { - "left": "issue:43525", - "right": "issue:44861", - "accept": false, - "reason": "Different config attribute bug vs tied-weight key handling; not the same defect." - }, - { - "left": "issue:31515", - "right": "issue:38617", - "accept": false, - "reason": "Both involve loading/import behavior, but one is slow checkpoint loading and the other is an ImportError; different root causes." - }, - { - "left": "issue:43572", - "right": "issue:44861", - "accept": false, - "reason": "Missing pad_token_idx in a model config is unrelated to tied-weight key crashes." - }, - { - "left": "issue:43066", - "right": "issue:44998", - "accept": false, - "reason": "Tokenizer decoder type issue vs off-topic issue; no overlap in underlying bug." - }, - { - "left": "issue:41669", - "right": "issue:44998", - "accept": false, - "reason": "Import-star performance problem is unrelated to the off-topic issue." - }, - { - "left": "issue:44998", - "right": "issue:46218", - "accept": false, - "reason": "Off-topic issue vs module import failure; not a duplicate set." - }, - { - "left": "issue:44230", - "right": "issue:44623", - "accept": false, - "reason": "FP8 inference support request is distinct from missing processor.save_pretrained files." - }, - { - "left": "issue:30333", - "right": "issue:31515", - "accept": false, - "reason": "MLflow job-status reporting bug vs slow from_pretrained checkpoint loading; different code paths." - }, - { - "left": "issue:31515", - "right": "issue:36296", - "accept": false, - "reason": "Slow checkpoint loading and tensor-parallel training bug are unrelated." - }, - { - "left": "issue:44230", - "right": "issue:45310", - "accept": false, - "reason": "Different Qwen3.5/MoE-related problems: fp8 support request vs from_pretrained error." - }, - { - "left": "issue:31515", - "right": "issue:33453", - "accept": false, - "reason": "Both mention loading, but one is checkpoint speed and the other tokenizer loading regression; not the same bug." - }, - { - "left": "issue:43298", - "right": "issue:44230", - "accept": false, - "reason": "Continuous batching audio support vs FP8 Qwen3-VL/MoE inference support are different feature gaps." - }, { "left": "issue:43717", "right": "issue:46144", "accept": false, - "reason": "Mamba init_weights behavior and RoFormer attention-interface usage are separate implementation issues." + "reason": "Different problems in different model components: init-weights behavior for Mamba/Mamba-2 vs RoFormer attention not using the attention interface." }, { "left": "issue:42936", "right": "issue:46018", "accept": false, - "reason": "Mask2Former ignore_value bug is unrelated to DeepSeek-V4 shared-expert gating." + "reason": "Mask2former ignore_value handling and DeepSeek-V4 shared-expert gating are unrelated code paths and symptoms." }, { "left": "issue:43668", "right": "issue:44031", "accept": false, - "reason": "Type-hint problem in ModernBERTConfig vs tokenizer regex warnings; different defects." + "reason": "A config type-hint issue in ModernBERT is not the same as a tokenizer regex warning regression affecting all tokenizers." }, { "left": "issue:42936", "right": "issue:43240", "accept": false, - "reason": "Mask2Former ignore_value handling and kwargs not reaching loss calculation are not the same bug." + "reason": "Both mention loss-related plumbing, but one is Mask2former ignore_value usage and the other is generic kwargs propagation to a loss function." }, { "left": "issue:43240", "right": "issue:46144", "accept": false, - "reason": "Loss-function kwargs propagation is unrelated to RoFormer attention implementation." + "reason": "Loss-function kwargs forwarding and RoFormer attention-interface usage are different subsystems and failure modes." }, { "left": "issue:42936", "right": "issue:44031", "accept": false, - "reason": "Mask2Former ignore_value issue does not match tokenizer warning behavior." + "reason": "Mask2former loss-argument handling is unrelated to a tokenizer regex warning after version 4.57.3." }, { "left": "issue:42936", "right": "issue:46144", "accept": false, - "reason": "Mask2Former ignore_value bug is unrelated to RoFormer attention interface usage." + "reason": "Unrelated model areas: Mask2former loss handling vs RoFormer attention implementation." }, { "left": "issue:42936", "right": "issue:43813", "accept": false, - "reason": "Mask2Former label handling issue and a quantization typo are distinct." + "reason": "Mask2former ignore_value usage and a quantization-attribute typo are not the same underlying bug." }, { "left": "issue:43240", "right": "issue:43668", "accept": false, - "reason": "Loss kwargs propagation bug is unrelated to a type-hint mismatch." + "reason": "Generic loss kwargs propagation and a ModernBERT type-hint correction are distinct, with no shared code path." }, { "left": "issue:43668", "right": "issue:43813", "accept": false, - "reason": "Type-hint issue and quantization typo do not share an underlying code defect." + "reason": "A type-hint fix and a typo in a quantization symbol are different kinds of issues in different areas." }, { "left": "issue:37934", "right": "issue:43668", "accept": false, - "reason": "Question about Llama4TextL2Norm behavior is unrelated to ModernBERTConfig typing." + "reason": "Both touch normalization-related terminology, but one questions Llama4TextL2Norm semantics while the other is just a ModernBERT config annotation bug." }, { "left": "issue:42936", "right": "issue:43668", "accept": false, - "reason": "Mask2Former ignore_value bug and ModernBERTConfig type hint issue are unrelated." + "reason": "Mask2former ignore_value logic and ModernBERT norm_eps typing are unrelated model bugs." } ] }, @@ -17322,12 +17207,12 @@ "issue:10519|issue:11005", "issue:11005|issue:11307" ], - "prepared_review_unit_hash": "f935840e8551ab16e8c9596d827d93f7869fc21944720740dbb049e80ba46c86", + "prepared_review_unit_hash": "a78dc1631eeeb3b190fc7ba626564ae69140d9ea1ad1bad2e6d62eb3f7b6e2ba", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2873, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -17335,7 +17220,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 2873, + "serialized_chars": 2876, "estimated_input_tokens": 719, "estimated_eval_tokens": 1694 }, @@ -17344,30 +17229,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These three issues are unrelated: one requests truncation from the beginning in tokenization, one asks for ReduceLROnPlateau-like learning-rate scheduling behavior, and one asks for start/end word offsets in Wav2Vec2. None look like duplicates or the same underlying bug/change.", - "confidence": 0.98, - "canonical_issue_reason": "No canonical issue: the cluster does not share a single topic or defect; the issues span tokenization, optimization/scheduling, and speech alignment.", + "summary": "The cluster does not contain duplicates: all three issues ask for different features in different parts of the library.", + "confidence": 0.97, + "canonical_issue_reason": "No canonical issue fits well because the issues are unrelated: tokenizer truncation behavior, ReduceLROnPlateau-like scheduling, and Wav2Vec2 word time offsets.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #10519 is the strongest standalone issue because it is the most concrete, narrowly scoped feature request with a clear API change request and good specificity.", + "best_issue_reason": "None of the issues is a strong cluster representative since they do not describe the same underlying bug or change.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:10519", "right": "issue:11307", "accept": false, - "reason": "Different subsystems and requests: token truncation behavior vs Wav2Vec2 word offsets. No common underlying bug or change." + "reason": "Different subsystems and requests: truncation-from-beginning for tokenization vs word boundary timestamps for Wav2Vec2." }, { "left": "issue:10519", "right": "issue:11005", "accept": false, - "reason": "One is about truncation direction in preprocessing; the other is about learning-rate scheduler functionality. They are unrelated." + "reason": "Unrelated functionality: tokenizer truncation options vs a ReduceLROnPlateau-like scheduler feature." }, { "left": "issue:11005", "right": "issue:11307", "accept": false, - "reason": "ReduceLROnPlateau-like scheduling and Wav2Vec2 word timestamp offsets are distinct feature requests with no shared code-path problem." + "reason": "Separate feature requests in different model areas with no shared concrete code-path or bug." } ] }, @@ -17399,12 +17284,12 @@ "issue:13244|issue:16618", "issue:15980|issue:16618" ], - "prepared_review_unit_hash": "d2cfbb3ec39cebf37c5c5a405a8227edda1ca8890bff373a0bd7180793a74ed0", + "prepared_review_unit_hash": "688441b0425578ffed0753a2c40d4511b5e6ead1097a29d8bd154b018ab6deb0", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2477, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -17412,7 +17297,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 2, - "serialized_chars": 2477, + "serialized_chars": 2478, "estimated_input_tokens": 620, "estimated_eval_tokens": 1496 }, @@ -17421,24 +17306,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The three issues do not appear to be duplicates of one another: one is about Tapas tokenization parity, one about a private-model auth/error-message problem, and one about failing to load bert-base-uncased. The only mild overlap is around model loading, but the specific failure modes are different.", - "confidence": 0.92, - "canonical_issue_reason": "Issue 16618 is the broadest model-loading report and the most plausible umbrella candidate, but it still does not match the other issues closely.", + "summary": "These three issues do not appear to be duplicates of one another: one is about Tapas tokenization mismatch, one about a private-model download/auth error, and one about failing to load bert-base-uncased. The only commonality is generic model-loading context, not the same underlying bug.", + "confidence": 0.96, + "canonical_issue_reason": "If forced to choose a representative issue, #16618 is the broadest and most actively discussed model-loading report, but it still does not subsume the other two topics.", "canonical_pr_reason": null, - "best_issue_reason": "16618 is the most central/general of the three and has the highest discussion activity, so it is the best representative issue if one must be chosen.", + "best_issue_reason": "#16618 is the best single issue to keep as the cluster representative because it is the most generic and most discussed of the three, though it should not be treated as a true duplicate target for the others.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:13244", "right": "issue:16618", "accept": false, - "reason": "Tokenization parity for Tapas is unrelated to a model-loading failure for bert-base-uncased." + "reason": "Tapas tokenization behavior is a different subsystem and bug from a generic bert-base-uncased loading failure; no shared code path is evident." }, { "left": "issue:15980", "right": "issue:16618", "accept": false, - "reason": "Private-model unauthorized-download messaging is a different bug than failing to load a public model." + "reason": "The private-model unauthorized download error is an auth/error-messaging problem, while #16618 is a model loading failure for a public model; these do not look like the same bug." } ] }, @@ -17468,12 +17353,12 @@ "soft_pairs": [ "issue:26350|issue:33260" ], - "prepared_review_unit_hash": "175ff552e918999ba673e3e8cc8ca8fa2fcf989f1194831d41ea95447192364a", + "prepared_review_unit_hash": "2c4f02c24ae771a174e046bec7589c8c8d77fe3bc3f3b475b0536e09e9dd5773", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -17481,7 +17366,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1589, + "serialized_chars": 1590, "estimated_input_tokens": 398, "estimated_eval_tokens": 1052 }, @@ -17490,18 +17375,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate feature-request issues about different capabilities: Flash Attention 2 support vs GGUF support. They are related only by being community contribution requests, not by the same underlying change or bug.", - "confidence": 0.97, - "canonical_issue_reason": null, + "summary": "These are two separate community contribution issues about adding support for different technologies: Flash Attention 2 vs GGUF. They share a template-like title but not the same underlying bug or change request, so they should not be clustered as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Neither issue is a duplicate of the other; they request different support additions for different features, so there is no clear canonical issue within this pair.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "If one must be chosen for the cluster, #26350 is slightly older and more broadly discussed, but it is not a better duplicate representative because the issues are unrelated.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:26350", "right": "issue:33260", "accept": false, - "reason": "Different requested features for different code paths and architectures; they do not describe the same underlying bug or change, so they should not be merged." + "reason": "Same issue template, but different feature areas and different requested changes; not the same underlying bug or enhancement." } ] }, @@ -17560,7 +17445,7 @@ "issue:44370|issue:46226", "issue:46222|issue:46226" ], - "prepared_review_unit_hash": "504eabf919691c09f054042f3795f43b00e27202c547ac12e19fd9077786d06d", + "prepared_review_unit_hash": "4700cac3a543216ab8e79ab1e4b6f909bb14ce3f931852f8a89103b9c5cad8a6", "component_budget": { "node_count": 8, "item_count": 8, @@ -17582,165 +17467,165 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Mostly separate language-specific translation requests. I reject the language-mismatched pairs, and only accept the placeholder-title cluster plus the Thai pair where the visible metadata is highly similar enough to suggest the same underlying issue.", - "confidence": 0.84, - "canonical_issue_reason": "Issue 27088 is the most concrete open issue: it names a specific target language and has substantial discussion activity, unlike the generic placeholder-style entries.", + "summary": "This cluster mixes several distinct i18n translation requests (Turkish, Spanish, Romanian) with two separate exact-duplicate subgroups: the placeholder-title translation issues and the repeated Thai-titled issue. Only the exact matches should be merged.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 43834 is the earliest of the exact-duplicate placeholder-title translation requests, making it the best canonical representative for that duplicate subgroup.", "canonical_pr_reason": null, - "best_issue_reason": "27088 is the strongest overall issue candidate because it is specific, open, and clearly scoped, while the other issues are either different languages or generic placeholder entries.", + "best_issue_reason": "43834 is the most suitable issue to keep because it is the earliest, generic representative of the duplicated placeholder-title translation request set.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:27088", "right": "issue:44370", "accept": false, - "reason": "Both are translation requests, but they target different languages; the shared i18n theme is not enough to call them duplicates." + "reason": "Turkish translation request vs. generic placeholder-title issue; no evidence they describe the same underlying request." }, { "left": "issue:27088", "right": "issue:44365", "accept": false, - "reason": "Different language targets; there is no visible evidence they describe the same translation request." + "reason": "Turkish translation request vs. generic placeholder-title issue; only broad i18n similarity." }, { "left": "issue:27088", "right": "issue:43834", "accept": false, - "reason": "A specific Turkish request versus a generic placeholder-style title; the visible metadata does not support a duplicate call." + "reason": "Concrete Turkish docs translation request is not the same as the placeholder-title issue." }, { "left": "issue:27088", "right": "issue:38435", "accept": false, - "reason": "Turkish vs Romanian translation requests are distinct issues." + "reason": "Different language requests (Turkish vs Romanian); not the same issue." }, { "left": "issue:27088", "right": "issue:28936", "accept": false, - "reason": "Turkish vs Spanish translation requests are different language-specific tasks." + "reason": "Different language requests (Turkish vs Spanish); not duplicates." }, { "left": "issue:28936", "right": "issue:43834", "accept": false, - "reason": "Spanish-specific request versus a generic placeholder entry; not enough evidence of the same underlying issue." + "reason": "Spanish translation request vs. generic placeholder-title issue; insufficient evidence of same underlying issue." }, { "left": "issue:28936", "right": "issue:44365", "accept": false, - "reason": "Different language targets, so these should stay separate unless there is stronger evidence of duplication." + "reason": "Spanish translation request vs. generic placeholder-title issue; only topic overlap." }, { "left": "issue:28936", "right": "issue:38435", "accept": false, - "reason": "Spanish and Romanian translation requests are separate issues." + "reason": "Different language requests (Spanish vs Romanian); not the same underlying issue." }, { "left": "issue:28936", "right": "issue:44370", "accept": false, - "reason": "Different language targets and no visible sign of the same request." + "reason": "Spanish translation request vs. generic placeholder-title issue; not a concrete duplicate match." }, { "left": "issue:38435", "right": "issue:44370", "accept": false, - "reason": "Romanian-specific request versus a generic placeholder entry; the metadata is similar in form but not the same language request." + "reason": "Romanian translation request vs. placeholder-title issue; no direct duplication evidence." }, { "left": "issue:38435", "right": "issue:44365", "accept": false, - "reason": "Romanian-specific request versus a generic placeholder entry; not enough evidence to merge them." + "reason": "Romanian translation request vs. placeholder-title issue; not the same issue." }, { "left": "issue:38435", "right": "issue:43834", "accept": false, - "reason": "A concrete Romanian request and a generic placeholder title are not clearly the same issue." + "reason": "Romanian translation request vs. placeholder-title issue; only broad i18n overlap." }, { "left": "issue:38435", "right": "issue:46222", "accept": false, - "reason": "Different language and title content; the visible metadata does not suggest the same underlying request." + "reason": "Romanian docs translation request and Thai-titled issue are unrelated." }, { "left": "issue:38435", "right": "issue:46226", "accept": false, - "reason": "Different language and title content; not a safe duplicate pair." + "reason": "Romanian docs translation request and Thai-titled issue are unrelated." }, { "left": "issue:43834", "right": "issue:44370", "accept": true, - "reason": "Same placeholder-style title and very close body lengths, with a very high similarity score; these look like duplicate copies of the same template-style issue." + "reason": "Exact same placeholder title and near-identical body shape; likely duplicate issues for the same translation template." }, { "left": "issue:43834", "right": "issue:44365", "accept": true, - "reason": "Same placeholder title and highly similar metadata, including near-matching body lengths and strong similarity, which is enough to treat them as duplicates." + "reason": "Same placeholder title/body pattern; looks like a duplicate of the same translation template issue." }, { "left": "issue:43834", "right": "issue:46222", "accept": false, - "reason": "Generic placeholder-style issue versus a Thai-titled issue; the visible metadata does not indicate the same request." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; different content and no same-request evidence." }, { "left": "issue:43834", "right": "issue:46226", "accept": false, - "reason": "Generic placeholder-style issue versus a Thai-titled issue; not enough evidence that they are the same issue." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not the same underlying issue." }, { "left": "issue:44365", "right": "issue:44370", "accept": true, - "reason": "Identical placeholder title and very similar body lengths, plus a very high similarity score, make this look like the same issue duplicated." + "reason": "Exact same placeholder title and highly similar bodies; strong duplicate signal." }, { "left": "issue:44365", "right": "issue:46222", "accept": false, - "reason": "Placeholder-style entry versus a Thai-titled issue; the metadata points to different issue contents." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; no basis to merge." }, { "left": "issue:44365", "right": "issue:46226", "accept": false, - "reason": "Placeholder-style entry versus a Thai-titled issue; not the same underlying request." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not duplicates." }, { "left": "issue:44370", "right": "issue:46222", "accept": false, - "reason": "Placeholder-style issue and Thai-titled issue are different on the face of the visible metadata." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; unrelated request text." }, { "left": "issue:44370", "right": "issue:46226", "accept": false, - "reason": "Placeholder-style issue and Thai-titled issue are different on the face of the visible metadata." + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not the same issue." }, { "left": "issue:46222", "right": "issue:46226", "accept": true, - "reason": "Identical Thai title, matching body length, and a perfect similarity score provide strong evidence these are duplicate entries of the same issue." + "reason": "Identical Thai title and matching body size/activity; this looks like a direct duplicate." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The overall direction is reasonable, but a few soft-edge reasons overstate what is visible in the packet. We only have titles, lengths, and similarity scores; there is no body text or structure shown, so claims like \"near-identical body structure,\" \"effectively identical-looking content,\" or \"exact duplicate\" are stronger than the evidence supports. The safer grounding is that the three placeholder-title issues and the two Thai-titled issues are highly similar by metadata, not proven identical. Please tighten those verdict reasons to stay conservative." + "accept": true, + "feedback": "The cluster summary is grounded in the packet: the three placeholder-title issues form a likely duplicate subgroup with 43834 as the earliest representative, and 46222/46226 are an exact-title duplicate pair. The Turkish, Spanish, and Romanian issues are kept separate, and the rejected cross-language/cross-template pairs are conservatively argued. No clear overstatement stands out." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -17813,12 +17698,12 @@ "issue:43837|issue:45177", "issue:45322|issue:46133" ], - "prepared_review_unit_hash": "7337d998f7a1e85009273458affc2bcd72d48426d6c91c0179f458fd4e2ed7bb", + "prepared_review_unit_hash": "ab5b1787b5d2f98cd3c22a9cff89910dc09a82a92f7e2a7ea09f9a7ba7ff821b", "component_budget": { "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21655, + "serialized_chars": 21654, "estimated_input_tokens": 5414, "estimated_eval_tokens": 11084 }, @@ -17826,7 +17711,7 @@ "node_count": 16, "item_count": 16, "soft_pair_count": 36, - "serialized_chars": 21655, + "serialized_chars": 21654, "estimated_input_tokens": 5414, "estimated_eval_tokens": 11084 }, @@ -17835,228 +17720,222 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This is a loose topical cluster of model-add/support requests, not a true duplicate set. The items share request wording and subsystem area, but they target different models, modalities, and underlying implementations.", - "confidence": 0.95, - "canonical_issue_reason": "Issue 43837 is the most central representative of the cluster by similarity links, but it is still just one specific model request among many distinct ones.", + "summary": "These are all broad model-addition requests, but they target different concrete models, modalities, or architecture changes. They are related thematically, not duplicates, so no soft pair should be merged. No PRs are present.", + "confidence": 0.94, + "canonical_issue_reason": "issue:36879 is a clean, concrete representative of the cluster\u2019s general \u2018add model support\u2019 theme, but it does not subsume the others.", "canonical_pr_reason": null, - "best_issue_reason": "43837 is the best single issue to represent the cluster\u2019s recurring 'add model support' theme because it sits near the center of the similarity graph.", + "best_issue_reason": "issue:36879 is the best representative anchor: short, specific, and easy to recognize as a model-support request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:36817", "right": "issue:41211", "accept": false, - "reason": "Different requested models (EuroBERT vs DEIMv2); same request pattern, not the same change." + "reason": "Different model requests: EuroBert vs DEIMv2." }, { "left": "issue:36817", "right": "issue:43671", "accept": false, - "reason": "EuroBERT config support and Qwen3-TTS support are unrelated model additions." + "reason": "Different targets and modalities: EuroBert vs Qwen3-TTS." }, { "left": "issue:36817", "right": "issue:36879", "accept": false, - "reason": "RF-DETR and EuroBERT are distinct architectures with different integration work." + "reason": "Different models: EuroBert vs RF-DETR." }, { "left": "issue:36879", "right": "issue:41211", "accept": false, - "reason": "RF-DETR and DEIMv2 are separate model requests, not one underlying bug/change." + "reason": "Different detector/model requests: RF-DETR vs DEIMv2." }, { "left": "issue:36879", "right": "issue:43909", "accept": false, - "reason": "RF-DETR and LFM2.5 Audio are different model families and code paths." + "reason": "Different modality and code path: RF-DETR vs LFM2.5 Audio." }, { "left": "issue:36879", "right": "issue:44121", "accept": false, - "reason": "RF-DETR and circuit-sparsity/circuitgpt are unrelated additions." + "reason": "Different change type: model support vs sparse-transformer architecture request." }, { "left": "issue:36879", "right": "issue:43837", "accept": false, - "reason": "RF-DETR and Qwen3-ASR are distinct models; similarity is only at the request-template level." + "reason": "Different modalities: RF-DETR vs Qwen3-ASR." }, { "left": "issue:36879", "right": "issue:43671", "accept": false, - "reason": "RF-DETR and Qwen3-TTS are different concrete model integrations." + "reason": "Different modalities: RF-DETR vs Qwen3-TTS." }, { "left": "issue:37015", "right": "issue:43837", "accept": false, - "reason": "NeoBERT and Qwen3-ASR are separate model requests with different implementation needs." + "reason": "Different requests: NeoBERT vs Qwen3-ASR." }, { "left": "issue:37015", "right": "issue:44121", "accept": false, - "reason": "NeoBERT and circuit-sparsity/circuitgpt are unrelated model changes." + "reason": "Different underlying changes: NeoBERT support vs sparse-transformer support." }, { "left": "issue:37015", "right": "issue:43909", "accept": false, - "reason": "NeoBERT and LFM2.5 Audio target different model types and code paths." - }, - { - "left": "issue:36879", - "right": "issue:37015", - "accept": false, - "reason": "RF-DETR and NeoBERT are different model integrations, not duplicates." + "reason": "Different domains: NeoBERT vs audio-model support." }, { "left": "issue:37015", "right": "issue:42503", "accept": false, - "reason": "NeoBERT and ModernVBERT are distinct model additions despite similar naming." + "reason": "Both are BERT-like, but they request distinct models and likely distinct integration work." }, { "left": "issue:41115", "right": "issue:43837", "accept": false, - "reason": "MiniCPM3 architecture support and Qwen3-ASR support are different model requests." + "reason": "Different modalities: MiniCPM3 architecture vs ASR support." }, { "left": "issue:41115", "right": "issue:44121", "accept": false, - "reason": "MiniCPM3 and circuit-sparsity/circuitgpt are not the same underlying feature." + "reason": "Different model families and changes: MiniCPM3 vs sparse transformer." }, { "left": "issue:41115", "right": "issue:42503", "accept": false, - "reason": "MiniCPM3 and ModernVBERT are separate model-add tasks." + "reason": "Different model requests: MiniCPM3 vs ModernVBERT." }, { "left": "issue:41115", "right": "issue:45027", "accept": false, - "reason": "MiniCPM3 and Voxtral-4B-TTS are distinct models and implementation paths." + "reason": "Different modalities: MiniCPM3 vs Voxtral TTS." }, { "left": "issue:41211", "right": "issue:43671", "accept": false, - "reason": "DEIMv2 and Qwen3-TTS are different model requests." + "reason": "Different model families and tasks: DEIMv2 vs Qwen3-TTS." }, { "left": "issue:42503", "right": "issue:43837", "accept": false, - "reason": "ModernVBERT models and Qwen3-ASR are unrelated model integrations." + "reason": "Different tasks/modalities: ModernVBERT vs Qwen3-ASR." }, { "left": "issue:42503", "right": "issue:43671", "accept": false, - "reason": "ModernVBERT and Qwen3-TTS are different concrete additions." + "reason": "Different tasks/modalities: ModernVBERT vs Qwen3-TTS." }, { "left": "issue:42503", "right": "issue:42738", "accept": false, - "reason": "ModernVBERT models and BERT-like RoPE support are related only broadly, not the same change." + "reason": "Related BERT theme, but one is a generic RoPE request and the other is a specific model request." }, { "left": "issue:42738", "right": "issue:45027", "accept": false, - "reason": "BERT-like RoPE support and Voxtral-4B-TTS are different requests." + "reason": "Unrelated requests: generic BERT-like RoPE support vs Voxtral TTS." }, { "left": "issue:43671", "right": "issue:43837", "accept": false, - "reason": "Qwen3-TTS and Qwen3-ASR are closely related by name but still different model integrations and code paths." + "reason": "Related audio family, but TTS and ASR are distinct model/code paths." }, { "left": "issue:43671", "right": "issue:45027", "accept": false, - "reason": "Qwen3-TTS and Voxtral-4B-TTS are separate audio model support requests." + "reason": "Both are TTS-related, but they are different models and integration work." }, { "left": "issue:43671", "right": "issue:45177", "accept": false, - "reason": "Qwen3-TTS and DeepSeek-OCR2 are unrelated model additions." + "reason": "Different modalities: TTS support vs OCR2 support." }, { "left": "issue:43671", "right": "issue:44121", "accept": false, - "reason": "Qwen3-TTS and circuit-sparsity/circuitgpt are different underlying changes." + "reason": "Different change types: TTS model support vs sparse-transformer model request." }, { "left": "issue:43837", "right": "issue:45322", "accept": false, - "reason": "Qwen3-ASR and EUPE are different model requests with no shared code path." + "reason": "Different modalities: ASR support vs perception-encoder support." }, { "left": "issue:43837", "right": "issue:43909", "accept": false, - "reason": "Qwen3-ASR and LFM2.5 Audio are separate model integrations." + "reason": "Related audio area, but distinct models and integration paths." }, { "left": "issue:43909", "right": "issue:44205", "accept": false, - "reason": "LFM2.5 Audio and SAM3-LiteText are different model additions." + "reason": "Different domains: audio model support vs SAM3-LiteText." }, { "left": "issue:43671", "right": "issue:43909", "accept": false, - "reason": "Qwen3-TTS and LFM2.5 Audio are unrelated audio model support requests." + "reason": "Different audio-model requests: Qwen3-TTS vs LFM2.5 Audio." }, { "left": "issue:44121", "right": "issue:45322", "accept": false, - "reason": "Weight-sparse transformer support and EUPE are different model proposals." + "reason": "Unrelated model families: sparse transformer vs perception encoder." }, { "left": "issue:43837", "right": "issue:44121", "accept": false, - "reason": "Qwen3-ASR and circuit-sparsity/circuitgpt are unrelated changes." + "reason": "Different requests: ASR support vs sparse-transformer support." }, { "left": "issue:43837", "right": "issue:45027", "accept": false, - "reason": "Qwen3-ASR and Voxtral-4B-TTS are different model support requests." + "reason": "Both audio-related, but ASR and TTS are distinct changes." }, { "left": "issue:37015", "right": "issue:45027", "accept": false, - "reason": "NeoBERT and Voxtral-4B-TTS are distinct model integrations." + "reason": "Different domains: NeoBERT support vs Voxtral TTS support." }, { "left": "issue:43837", "right": "issue:45177", "accept": false, - "reason": "Qwen3-ASR and DeepSeek-OCR2 are separate model requests." + "reason": "Different tasks/modalities: ASR support vs OCR2 support." }, { "left": "issue:45322", "right": "issue:46133", "accept": false, - "reason": "EUPE and TIPSv2 are different research models; same broad theme, not the same underlying change." + "reason": "Different vision-language model requests: EUPE vs TIPSv2." } ] }, @@ -18086,12 +17965,12 @@ "soft_pairs": [ "issue:38740|issue:46443" ], - "prepared_review_unit_hash": "0373b5aba99d14bf0487cc28b63c144c3ef3f18b1e19a709744bf9952eec8b99", + "prepared_review_unit_hash": "8e212551ab9be4ffb6841a92be4b0511521a70b31d57eb0458e8246c631581d4", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1506, + "serialized_chars": 1505, "estimated_input_tokens": 377, "estimated_eval_tokens": 1010 }, @@ -18099,7 +17978,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1506, + "serialized_chars": 1505, "estimated_input_tokens": 377, "estimated_eval_tokens": 1010 }, @@ -18108,18 +17987,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are unrelated: one is a docs request to add Pruna as an optimization framework, the other is a Vision RoPE refactor. The soft similarity signal is not enough to treat them as duplicates.", - "confidence": 0.98, - "canonical_issue_reason": null, + "summary": "These two issues are not duplicates: one is a docs request to add Pruna as an optimization framework, while the other is a Vision RoPE refactor. They do not describe the same underlying bug or change.", + "confidence": 0.99, + "canonical_issue_reason": "No canonical issue: the two issues are unrelated in scope and purpose (documentation request vs model internals refactor).", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Neither issue is a good canonical duplicate target because they concern different subsystems and different kinds of work.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:38740", "right": "issue:46443", "accept": false, - "reason": "Different subjects and problem domains (documentation/framework listing vs model architecture refactor); no shared targets, filenames, or concrete code-path overlap." + "reason": "Different topics and change types; the first is a docs/optimization-framework request, the second is a Vision RoPE code refactor." } ] }, @@ -18149,7 +18028,7 @@ "soft_pairs": [ "issue:41084|issue:43037" ], - "prepared_review_unit_hash": "8b2f46f37826fb02b844d8c8dbb1930e8c4af57b82d9a95f02e6208565407286", + "prepared_review_unit_hash": "d8f480ddb7b81dad3662f6b227689f6f78ad257ff090d169f89813905215ffc7", "component_budget": { "node_count": 2, "item_count": 2, @@ -18171,18 +18050,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two issues do not look like duplicates: one is a broad request about block decoding, while the other is a specific model-support request for DeepSeek v3.2.", - "confidence": 0.97, - "canonical_issue_reason": "Issue #43037 is the better anchor because it is a concrete, user-visible feature request with a specific target; #41084 is more generic and less clearly scoped.", + "summary": "These two issues are not duplicates: one is a generic \"Set Block Decoding\" request, while the other is specifically about adding DeepSeek v3.2 support. They describe different user needs and likely different code paths.", + "confidence": 0.96, + "canonical_issue_reason": null, "canonical_pr_reason": null, - "best_issue_reason": "#43037 is the more suitable standalone issue since it names a precise capability request, whereas #41084 is too broad to serve well as the canonical representative.", + "best_issue_reason": "Issue #43037 is the more concrete and actionable request; #41084 is too vague to serve as a strong canonical representative.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:41084", "right": "issue:43037", "accept": false, - "reason": "Different underlying requests: block decoding behavior vs support for a specific model/version. They do not appear to describe the same bug or change." + "reason": "Different underlying topics: block decoding behavior vs model/version support. No clear shared bug or change scope." } ] }, @@ -18212,7 +18091,7 @@ "soft_pairs": [ "issue:43598|issue:43600" ], - "prepared_review_unit_hash": "5a98482684c443f2b7dde8e1193895860c94fc93550e18a0fe490aadba445caf", + "prepared_review_unit_hash": "909006205d4b5a0aa0ed6950914cf26847e07222a740124f4a84b1005146d76a", "component_budget": { "node_count": 2, "item_count": 2, @@ -18234,18 +18113,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are related only at a broad Trainer-customization level, but they describe different problems: one is about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They should not be treated as duplicates.", - "confidence": 0.93, - "canonical_issue_reason": "Issue #43598 is the more concrete, code-path-specific request and reads like a focused Trainer behavior change rather than a general instrumentation concern.", + "summary": "These are related to Trainer customization, but they describe different problems: one about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They should not be deduplicated.", + "confidence": 0.91, + "canonical_issue_reason": "Issue #43598 is the broader, more central Trainer customizability request; #43600 is a separate FLOPS-estimation concern rather than the same bug.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #43598 is the strongest standalone issue because it names a specific Trainer knob and a clear customization problem, making it more actionable and likely to anchor related discussion.", + "best_issue_reason": "#43598 is the better canonical issue because it frames the main Trainer customization limitation more generally.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43598", "right": "issue:43600", "accept": false, - "reason": "Different underlying concerns: `remove_unused_column` customizability versus FLOPS estimation for custom trainers. They share the Trainer/custom trainer context but not the same bug or change." + "reason": "Both mention custom trainers/Trainer internals, but one is about `remove_unused_column` customizability and the other is about estimating FLOPS. Different code paths and different fixes." } ] }, @@ -18275,7 +18154,7 @@ "soft_pairs": [ "issue:43684|issue:44405" ], - "prepared_review_unit_hash": "e5a1dcdfdb41841739338f331d34c7d749c1fdd1de48acd6a8e5ab42b363be78", + "prepared_review_unit_hash": "5fa4b56f9e4eb15d356ef9cb5dec3bd0f7f7fba3a8d0e5db7dbfb4e058c9ec33", "component_budget": { "node_count": 2, "item_count": 2, @@ -18297,18 +18176,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: one asks for Qwen3-Omni auto-registration for generation/model loading, while the other asks for Qwen3.5 sequence-classification support. Different model families and different AutoModel targets.", - "confidence": 0.94, - "canonical_issue_reason": "43684 is the broader, more general support request and is the better representative of the cluster, but it is still distinct from 44405.", + "summary": "These issues are related only at a very broad level (adding AutoModel support for Qwen variants), but they target different model families and different auto-class behaviors, so they are not duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #43684 is the broader, more foundational registration request (AutoModel and AutoModelForConditionalGeneration for Qwen3-Omni) and is the better canonical issue if one must be chosen.", "canonical_pr_reason": null, - "best_issue_reason": "43684 is the closest cluster representative because it concerns core AutoModel/AutoModelForConditionalGeneration registration rather than a specific task head.", + "best_issue_reason": "#43684 is the most central issue in this pair because it asks for core model registration rather than a narrower task-specific classifier mapping.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43684", "right": "issue:44405", "accept": false, - "reason": "Different underlying changes: Qwen3-Omni registration vs Qwen3.5 sequence-classification support. Not the same bug or patch." + "reason": "Different underlying request: Qwen3-Omni model registration for AutoModel/AutoModelForConditionalGeneration vs. Qwen3.5 sequence-classification support for AutoModelForSequenceClassification. Different model family and code path." } ] }, @@ -18345,70 +18224,70 @@ "issue:43845|issue:43859", "issue:43846|issue:43859" ], - "prepared_review_unit_hash": "887753543e19f593e61bcea3313ce5c5ff94f7e55c61e3956b799f3061e887ef", + "prepared_review_unit_hash": "11710777dcc152416c6a73c67cdc30a73e2ac49e49e39bbf858258c03ea27f21", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4263, - "estimated_input_tokens": 1066, - "estimated_eval_tokens": 2388 + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 4263, - "estimated_input_tokens": 1066, - "estimated_eval_tokens": 2388 + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The cluster mostly contains low-information issues with repeated title text. Three issues share the same title, but there isn\u2019t enough concrete description to confirm they are the same bug; the remaining issue ('Fett') looks unrelated.", - "confidence": 0.43, - "canonical_issue_reason": "Issue 43845 is the best canonical anchor: it is one of the repeated 'huggingface' reports, appears earliest among that group, and has slightly more body content than 43859.", + "summary": "One low-signal cluster: three issues share the exact title \"huggingface\" and look related by repetition, while \"Fett\" appears unrelated. Evidence is too sparse to confidently merge any pair.", + "confidence": 0.41, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface\" issues and is the most reasonable representative of that subgroup; 43835 has a different title and stands apart.", "canonical_pr_reason": null, - "best_issue_reason": "43845 is the strongest representative of the cluster because it matches the repeated title pattern and is the earliest/most substantial of the similar issues.", + "best_issue_reason": "43845 is the best issue to keep as the cluster representative because it is the earliest identical-title report and there is no richer content to distinguish the duplicates.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:43835", "right": "issue:43845", "accept": false, - "reason": "Different titles and no shared concrete problem details; not enough evidence that these are the same underlying issue." + "reason": "Different titles and no substantive body detail tying \"Fett\" to the repeated \"huggingface\" reports." }, { "left": "issue:43835", "right": "issue:43846", "accept": false, - "reason": "Different titles and no substantive description tying them to the same bug." + "reason": "Different titles and no evidence they describe the same underlying problem." }, { "left": "issue:43835", "right": "issue:43859", "accept": false, - "reason": "Different titles and no concrete overlap beyond being in the same cluster." + "reason": "No meaningful overlap beyond being closed issues; the titles do not suggest the same bug." }, { "left": "issue:43845", "right": "issue:43846", "accept": false, - "reason": "Same short title, but no meaningful description or issue target to confirm they are the same bug; too little evidence to merge." + "reason": "These look duplicate-like because the titles match, but there is no issue description detail to confirm the same underlying bug or change." }, { "left": "issue:43845", "right": "issue:43859", "accept": false, - "reason": "Shared title text alone is insufficient; the bodies are too sparse to verify the same underlying issue." + "reason": "Exact same title, but the reports are too content-poor to verify they refer to the same concrete issue." }, { "left": "issue:43846", "right": "issue:43859", "accept": false, - "reason": "Shared title text alone is insufficient; no concrete bug details are available to justify treating them as duplicates." + "reason": "Exact same title, yet the available evidence is insufficient to assert a shared underlying bug/change." } ] }, @@ -18449,88 +18328,88 @@ "issue:44572|issue:44937", "issue:44573|issue:44937" ], - "prepared_review_unit_hash": "d0fdf18387fd90044b90771c308833b052ce460b3dd9648d1c869382bea8bf62", + "prepared_review_unit_hash": "fe89be851ada82bfb850dc24de646bfcc731f36b386f1ecc855ec9cde2e19115", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5791, - "estimated_input_tokens": 1448, - "estimated_eval_tokens": 3152 + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 9, - "serialized_chars": 5791, - "estimated_input_tokens": 1448, - "estimated_eval_tokens": 3152 + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "No convincing duplicate cluster here: the items are low-information spam/doc-suggestion issues with unrelated titles and no shared issue targets or filenames. Treat them as separate reports.", + "summary": "This cluster is mostly spam/placeholder issues with no evidence of shared bug reports. Issue 44418 is the only one with a meaningful title, so it is the best anchor, but none of the pairings look like true duplicates.", "confidence": 0.94, - "canonical_issue_reason": null, + "canonical_issue_reason": "issue:44418 is the only non-spam, descriptive issue in the set, so it is the best canonical anchor among otherwise low-signal reports.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "issue:44418 is the most substantive and least noisy issue here; the others are generic spam or similarly uninformative.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:44371", "right": "issue:44572", "accept": false, - "reason": "Both are generic spam-like issues, but there is no evidence they describe the same underlying problem." + "reason": "Both are generic '' placeholders with no concrete shared bug/change to deduplicate." }, { "left": "issue:44371", "right": "issue:44573", "accept": false, - "reason": "Same as above: identical low-information formatting is not enough to establish a duplicate." + "reason": "Both are generic '' placeholders with no concrete shared bug/change to deduplicate." }, { "left": "issue:44371", "right": "issue:44418", "accept": false, - "reason": "One is spam, the other is a documentation suggestion; these do not look like the same issue." + "reason": "One is spam, the other is a documentation suggestion; the titles do not indicate the same underlying issue." }, { "left": "issue:44371", "right": "issue:44937", "accept": false, - "reason": "Completely different titles and no shared technical context." + "reason": "Generic spam is not the same underlying issue as the 'Google LLC' title, which is a different topic entirely." }, { "left": "issue:44418", "right": "issue:44572", "accept": false, - "reason": "A documentation enhancement request and a spam placeholder are unrelated." + "reason": "A documentation enhancement request does not match a generic spam placeholder." }, { "left": "issue:44418", "right": "issue:44573", "accept": false, - "reason": "No shared underlying bug or change is apparent." + "reason": "A documentation enhancement request does not match a generic spam placeholder." }, { "left": "issue:44572", "right": "issue:44573", "accept": false, - "reason": "Both are spam placeholders, but there is no concrete evidence they refer to the same report." + "reason": "Both are spam placeholders, but there is no evidence they describe the same concrete bug or change." }, { "left": "issue:44572", "right": "issue:44937", "accept": false, - "reason": "Generic spam versus a specific quoted-title issue; not the same underlying topic." + "reason": "Generic spam and a 'Google LLC' topic are unrelated on the available evidence." }, { "left": "issue:44573", "right": "issue:44937", "accept": false, - "reason": "No substantive overlap beyond being unrelated closed issues." + "reason": "Generic spam and a 'Google LLC' topic are unrelated on the available evidence." } ] }, @@ -18560,12 +18439,12 @@ "soft_pairs": [ "issue:45145|issue:45151" ], - "prepared_review_unit_hash": "90905a85d6e298d60b2eca35be83c1cc75273c1f923edab09fa0299279dba4f9", + "prepared_review_unit_hash": "a1817adf62117243a2d32e5f94bf25a32fe2c69ea6edf7548c18fc59c763e87f", "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 }, @@ -18573,7 +18452,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1646, + "serialized_chars": 1645, "estimated_input_tokens": 412, "estimated_eval_tokens": 1080 }, @@ -18582,28 +18461,28 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are likely duplicates: they have identical titles and a soft-similarity signal, with #45145 appearing to be the better canonical target because it was opened first and already has an inbound reference. The evidence does not support a stronger claim than likely duplication.", - "confidence": 0.86, - "canonical_issue_reason": "#45145 is the earlier issue and has an inbound reference, so it is the better canonical duplicate target.", + "summary": "The two closed issues look highly similar, but the packet only supports a cautious association based on identical titles and soft-similarity signals. #45145 is the better representative because it is earlier and has the only inbound reference, but the evidence is not strong enough to confidently assert a duplicate relationship.", + "confidence": 0.74, + "canonical_issue_reason": "#45145 is the earlier issue and has the only inbound reference, so it is the best representative of the pair if they are ultimately treated as related. The packet does not include bodies, so this should be framed as a representative choice rather than a proven duplicate determination.", "canonical_pr_reason": null, - "best_issue_reason": "#45145 is the best representative issue because it is older and already referenced inbound.", + "best_issue_reason": "#45145 is the best issue representative because it is earlier and has an inbound reference, while the available evidence does not show any substantive difference from #45151. This is a conservative selection rather than a strong duplicate claim.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45145", "right": "issue:45151", - "accept": true, - "reason": "Identical titles and the soft-similarity signal make these look like the same underlying issue report, though the evidence is limited." + "accept": false, + "reason": "The identical title and soft-similarity signal suggest a possible duplicate, but the packet does not provide issue bodies and the pair is marked deterministic_accept:false. That is not enough evidence to confidently accept the edge as the same underlying issue." } ] }, "evaluator_result": { "accept": false, - "feedback": "The duplicate/canonical call is plausible, but the summary and soft-edge reason overstate the evidence. The packet only shows identical titles, similar body lengths, and a soft-similarity signal; it does not support claims of an \"exact same\" proposal or \"near-identical bodies.\" Keep the canonical preference for #45145 as the earlier issue with an inbound reference, but tone down the duplicate certainty and body-content claims." + "feedback": "The verdict is too strong for the evidence in the packet. We only have same title, similar body lengths, and a soft-similarity signal; the packet does not provide the issue bodies, so claims like \"near-identical content\" or \"clear duplicates\" are not fully grounded. Also, the soft pair is marked `deterministic_accept: false`, so accepting the duplicate edge is not conservative enough. The canonical-choice rationale can mention that #45145 is earlier and has an inbound reference, but it should not overstate substantiated duplication." }, "evaluator_used": true, "retried": true, - "accepted_nontrivial_soft_edge": true, + "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null }, @@ -18680,40 +18559,40 @@ "soft_pairs": [ "issue:45672|issue:45674" ], - "prepared_review_unit_hash": "0303233f2a9845a40240472897a189998045c2106fb0002a18d1df580e371f20", + "prepared_review_unit_hash": "0344bfe59ff04a353e88848e8f983e436f5fa7ad0ace78e7abca33021bda56ab", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1672, - "estimated_input_tokens": 418, - "estimated_eval_tokens": 1092 + "serialized_chars": 1673, + "estimated_input_tokens": 419, + "estimated_eval_tokens": 1094 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1672, - "estimated_input_tokens": 418, - "estimated_eval_tokens": 1092 + "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": "Both items are closed issues about 4-bit inference, but they describe different failure modes and code paths: one is a Gemma4 audio-module uint8/torch.finfo TypeError, the other is a BitsAndBytesConfig bug where llm_int8_skip_modules removes the default lm_head exclusion and triggers an AssertionError.", - "confidence": 0.95, - "canonical_issue_reason": "issue 45672 is a clean, specific bug report with a narrow symptom and context (Gemma4 audio modules, uint8 weights, NF4 inference), making it a good representative issue for its distinct failure mode.", + "summary": "These are two separate 4-bit inference bugs: one about torch.finfo() failing on uint8 weights in Gemma4 audio modules, the other about llm_int8_skip_modules wiping the default lm_head exclusion and triggering an AssertionError. They are related only at a broad BitsAndBytes/4-bit level, not the same underlying defect.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #45674 is the better representative because it describes a concrete configuration regression in BitsAndBytesConfig with a clearer failure mode and broader impact.", "canonical_pr_reason": null, - "best_issue_reason": "issue 45672 is the better canonical choice because it is more concrete and self-contained; issue 45674 is a separate configuration bug in a different part of the 4-bit inference flow.", + "best_issue_reason": "Issue #45674 is the strongest standalone issue in the set: it is specific, actionable, and maps to a distinct code-path problem rather than a model-specific symptom.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45672", "right": "issue:45674", "accept": false, - "reason": "Reject: these are not the same underlying bug. 45672 is a torch.finfo TypeError on uint8 weights in audio modules, while 45674 is an AssertionError caused by llm_int8_skip_modules clearing the default lm_head exclusion in BitsAndBytesConfig." + "reason": "Reject: the failures are different. #45672 is a Gemma4/audio-module torch.finfo() uint8-type issue, while #45674 is an llm_int8_skip_modules / lm_head exclusion regression causing AssertionError. Same subsystem, different bug." } ] }, @@ -18743,7 +18622,7 @@ "soft_pairs": [ "issue:45762|issue:45763" ], - "prepared_review_unit_hash": "f8ad3ab02578b611f651fc009cf3a76fafa9d90f3f06979335ebea64750231bb", + "prepared_review_unit_hash": "5ee06a725d5ffa0809c5e835c2dc67cb5cc297018db98101d007d69e63b52ed3", "component_budget": { "node_count": 2, "item_count": 2, @@ -18765,18 +18644,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are not duplicates; they are unrelated personal-name/date entries with no shared problem description, references, or evidence of the same underlying bug.", + "summary": "Two closed issues with unrelated personal-name titles and different dates; they do not describe the same bug or request.", "confidence": 0.98, - "canonical_issue_reason": null, + "canonical_issue_reason": "Issue 45762 is a reasonable canonical only by default; it has no body or technical context, so there is no substantive basis to prefer it over 45763.", "canonical_pr_reason": null, - "best_issue_reason": null, + "best_issue_reason": "Neither issue is a strong canonical bug report, but 45762 is marginally as good a representative as 45763 because both are empty, non-technical, and equally non-descriptive.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45762", "right": "issue:45763", "accept": false, - "reason": "Reject: these look like two separate, unrelated issues with only weak textual similarity in names/dates; there is no indication they report the same bug or request." + "reason": "The titles are different personal names with different dates and no shared technical details, explicit links, or overlapping behavior; they appear unrelated rather than duplicate." } ] }, @@ -18806,7 +18685,7 @@ "soft_pairs": [ "issue:45764|issue:45765" ], - "prepared_review_unit_hash": "a283a6752b28dbf5adc6ffe1095f850570a0e48426d7e362e38951b9ca4b9e37", + "prepared_review_unit_hash": "74fb955d415de602c7ee447936971f0274f873516dd94ae2e8b7d45a3db29880", "component_budget": { "node_count": 2, "item_count": 2, @@ -18828,8 +18707,8 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two closed issues are not duplicates: their titles are unrelated personal-name/date strings, with no body or linking evidence suggesting the same underlying problem. They look more like isolated placeholder/spam reports than a shared bug or task.", - "confidence": 0.98, + "summary": "The two closed issues are just name/date placeholders and do not describe a shared technical problem, so they should not be clustered as duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, @@ -18839,7 +18718,7 @@ "left": "issue:45764", "right": "issue:45765", "accept": false, - "reason": "No substantive overlap beyond both being sparse closed issues; the titles reference different people and different dates, with no shared codepath, description, or explicit linkage." + "reason": "No bug report, feature request, or code-path description is present; the titles only contain different personal names and dates, so there is no evidence they refer to the same underlying issue." } ] }, @@ -18869,40 +18748,40 @@ "soft_pairs": [ "issue:45810|issue:46295" ], - "prepared_review_unit_hash": "07f5dd908ff814bbbfeb574de0214ba502bd640508a5b70ed612d056a48d9aaf", + "prepared_review_unit_hash": "e7ed51b01c7d79a02b2d10d3b7410eec7d9196461c2ca1ccde215c243b887b31", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1541, - "estimated_input_tokens": 386, - "estimated_eval_tokens": 1028 + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 1541, - "estimated_input_tokens": 386, - "estimated_eval_tokens": 1028 + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These two open issues are unrelated: one requests a new Qwen3_5 token-classification/value-model class, while the other asks for FP8 training support. They do not describe the same underlying bug or feature.", - "confidence": 0.98, - "canonical_issue_reason": "Neither issue is a clear duplicate of the other; if forced to pick a representative, #45810 is the more specific code/API request, but there is no true canonical duplicate here.", + "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 touch different features and code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #46295 is the broader, more general feature request; #45810 is a narrow model-specific addition and does not match it.", "canonical_pr_reason": null, - "best_issue_reason": "Issue #45810 is the more concrete product request and is easier to route to the right model/API area; #46295 is a separate training-capability request.", + "best_issue_reason": "#46295 is the better representative issue for the cluster because it is the more general and potentially higher-impact request.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45810", "right": "issue:46295", "accept": false, - "reason": "Different requests in different areas: adding a Qwen token-classification model class vs. enabling FP8 training. No shared bug or implementation path." + "reason": "Different underlying requests: adding a Qwen3_5 token-classification/value-model class versus enabling FP8 training. No shared bug or concrete code-path fix." } ] }, @@ -18942,88 +18821,88 @@ "issue:45871|issue:45873", "issue:45872|issue:45873" ], - "prepared_review_unit_hash": "71df1866659621e88b88e4701a8eb15bb81b832b3ef456e51f5321799cdf7c99", + "prepared_review_unit_hash": "344a4eabb7fbfe3230d4be74f88bd95311db751bd83bb0b5c52ef183da2b059b", "component_budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, - "estimated_input_tokens": 1432, - "estimated_eval_tokens": 3120 + "serialized_chars": 5729, + "estimated_input_tokens": 1433, + "estimated_eval_tokens": 3122 }, "budget": { "node_count": 5, "item_count": 5, "soft_pair_count": 8, - "serialized_chars": 5726, - "estimated_input_tokens": 1432, - "estimated_eval_tokens": 3120 + "serialized_chars": 5729, + "estimated_input_tokens": 1433, + "estimated_eval_tokens": 3122 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is not one single duplicate set; it splits into two exact-duplicate issue pairs: 45870/45871 and 45872/45873. The dataset/sample issue 45869 is only loosely related by topic and should stay separate.", - "confidence": 0.89, - "canonical_issue_reason": "Issue 45872 is the strongest anchor because it is the more specific, better-substantiated request and has an exact duplicate in 45873. The cluster overall contains two separate duplicate subgroups, but 45872/45873 looks like the best representative pair.", + "summary": "There are two exact duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is thematically similar but appears to be a different dataset promotion topic, not the same underlying request/bug. No pull requests are present.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45870 is the best canonical issue because it is the earliest and slightly more active of the exact duplicate pair 45870/45871; 45871 is a straight duplicate, while 45869 and 45872/45873 are separate topics.", "canonical_pr_reason": null, - "best_issue_reason": "45872 is the best issue to keep as the representative issue because it is a clear duplicate pair with 45873 and is more specific than the shorter example/prediction pair.", + "best_issue_reason": "issue:45870 is the strongest representative issue in the cluster: it is an exact duplicate target, has a duplicate partner, and is a cleaner canonical anchor than the unrelated dataset promo issue.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:45869", "right": "issue:45870", "accept": false, - "reason": "Related financial/data theme, but one is a dataset sample and the other is an example/prediction request; not the same underlying issue." + "reason": "Related financial/DataForge wording, but one is a dataset promotion and the other is an example request; not the same underlying issue." }, { "left": "issue:45869", "right": "issue:45871", "accept": false, - "reason": "Dataset/sample request is distinct from the example/prediction request despite shared finance wording." + "reason": "Same as 45869/45870: topical overlap only, not a true duplicate." }, { "left": "issue:45870", "right": "issue:45871", "accept": true, - "reason": "Exact title match and highly similar metadata; these are the same issue request." + "reason": "Exact same title and effectively the same issue content; clear duplicate." }, { "left": "issue:45870", "right": "issue:45872", "accept": false, - "reason": "Both mention DataForge and finance, but one is an example/prediction request while the other is a scraping-service request; different change/issue." + "reason": "Both are DataForge/financial-themed, but they describe different requests (example prediction vs scraping service)." }, { "left": "issue:45870", "right": "issue:45873", "accept": false, - "reason": "Different underlying request: example/prediction versus scraping-service integration." + "reason": "Different request from 45870; similar surface wording but not the same underlying issue." }, { "left": "issue:45871", "right": "issue:45872", "accept": false, - "reason": "Same mismatch as 45870/45872: related topic, but not the same concrete request." + "reason": "Exact-duplicate partner of 45870, but not the same topic as the scraping-service issue." }, { "left": "issue:45871", "right": "issue:45873", "accept": false, - "reason": "Same mismatch as 45870/45873: different requested outcome." + "reason": "Different request from 45871; similar subsystem mention only." }, { "left": "issue:45872", "right": "issue:45873", "accept": true, - "reason": "Exact title match and matching issue framing; these are duplicates of the same request." + "reason": "Exact same title and body shape; clear duplicate." } ] }, "evaluator_result": { "accept": true, - "feedback": "Grounded overall. The two accepted duplicate pairs are well supported by identical titles and matching metadata, and the non-duplicate verdicts stay conservative given only soft similarity. The only minor concern is a bit of rhetorical overreach in the canonical issue reason, but not enough to overturn the cluster read." + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 have identical titles and matching body lengths, while 45869 is clearly a different dataset-promo topic. The canonical choice of 45870 is conservative and supported by earlier creation time plus slightly higher discussion activity. No PRs are present. Soft-edge accepts are limited to the two strongest duplicate-like pairs and do not overreach." }, "evaluator_used": true, "retried": false, @@ -19050,7 +18929,7 @@ "soft_pairs": [ "issue:46439|issue:46441" ], - "prepared_review_unit_hash": "33b7d7b7ee2ae9a625563abd39b783645ec50a13f60148af9925a9026ef629b9", + "prepared_review_unit_hash": "3c9fa5f880d5c6ad9484e7ae2d859b64e0e371e8aad335b8b994823703d04fa4", "component_budget": { "node_count": 2, "item_count": 2, @@ -19072,27 +18951,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two issues are plausibly related because both involve `cache_implementation=\"static\"` and a failure in the linear-attention path, but the available evidence is too thin to treat 46441 as a confirmed manifestation of 46439. With only title-level overlap and no shared targets, filenames, or body-confirmed root cause, the duplicate link should stay conservative.", - "confidence": 0.62, - "canonical_issue_reason": "46439 is the better canonical issue because it describes the broader defect class in static-cache early initialization and the linear-attention shape corruption, rather than a model-specific crash symptom.", + "summary": "These two issues are related by static caching and linear-attention, but they describe different failure modes: one is a shape corruption during `early_initialization`, the other is a `KeyError` for a missing `linear_attention` cache entry in Qwen3.5.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it states the broader underlying static-cache/linear-attention corruption and gives a more general bug description than the model-specific KeyError report.", "canonical_pr_reason": null, - "best_issue_reason": "46439 is the strongest issue to keep as the primary target since it captures the underlying code-path problem more generally than 46441.", + "best_issue_reason": "Issue 46439 is the strongest representative of the cluster: it is more general, describes the failure mechanism, and is more likely to map to the core code-path problem.", "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "issue:46439", "right": "issue:46441", "accept": false, - "reason": "Plausibly overlapping static-cache/linear-attention failures, but the record only provides soft similarity and no shared targets, filenames, or explicit root-cause confirmation. That is not enough to say 46441 is a specific manifestation of 46439." + "reason": "Reject: they involve the same subsystem, but the reported failures differ materially (`RuntimeError` from wrong shapes during early initialization vs `KeyError: 'linear_attention'` in Qwen3.5 static cache), so they do not clearly share the same underlying bug." } ] }, - "evaluator_result": { - "accept": false, - "feedback": "The verdict is too strong for the packet. The only evidence provided for the pair is `soft_similarity` with no shared targets or filenames, and the titles alone do not establish that 46441 is caused by the broader defect in 46439. Please keep this conservative: say they are plausibly related or overlapping, not that 46441 is a specific manifestation of 46439. The accept on the soft pair should be rejected or downgraded unless the bodies explicitly confirm the same root cause." - }, - "evaluator_used": true, - "retried": true, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, "accepted_nontrivial_soft_edge": false, "error_kind": null, "error_message": null @@ -19100,7 +18976,7 @@ { "label": "pull_request", "component_index": 1, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19108,7 +18984,7 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-39895-15", + "cluster_id": "cluster-39895-14", "nodes": [ "pull_request:39895", "pull_request:41855", @@ -19123,13 +18999,12 @@ "pull_request:45213", "pull_request:45655", "pull_request:46048", - "pull_request:46267", - "pull_request:46495" + "pull_request:46267" ], "soft_pairs": [ "pull_request:39895|pull_request:43424", - "pull_request:41855|pull_request:43424", "pull_request:41855|pull_request:43823", + "pull_request:41855|pull_request:43424", "pull_request:42668|pull_request:43823", "pull_request:43251|pull_request:43823", "pull_request:43251|pull_request:44827", @@ -19139,169 +19014,162 @@ "pull_request:43823|pull_request:43995", "pull_request:43995|pull_request:45213", "pull_request:43995|pull_request:44827", - "pull_request:44827|pull_request:46495", "pull_request:44827|pull_request:45170", "pull_request:44827|pull_request:46267", "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", "pull_request:44827|pull_request:45213", "pull_request:43823|pull_request:45213", - "pull_request:44827|pull_request:45655", "pull_request:43251|pull_request:43254" ], - "prepared_review_unit_hash": "52ceadf2a2caed291e3e44542023a6a0ef79d817251052c10fb1dbd40e6a7261", + "prepared_review_unit_hash": "c4cb72318b8deea24ca2d2844f50eba7649156d89c90a30192515568a359f5bd", "component_budget": { - "node_count": 15, - "item_count": 15, - "soft_pair_count": 20, - "serialized_chars": 40462, - "estimated_input_tokens": 10116, - "estimated_eval_tokens": 20488 + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 }, "budget": { - "node_count": 15, - "item_count": 15, - "soft_pair_count": 20, - "serialized_chars": 40462, - "estimated_input_tokens": 10116, - "estimated_eval_tokens": 20488 + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is almost entirely unrelated open PRs; the only true duplicate-like pair is the two `fixed_cross_entropy` fixes. PR #43251 is the best canonical representative because it directly and more completely addresses the shared loss-kwargs bug and links the tracking issue.", - "confidence": 0.91, + "summary": "This cluster is mostly a grab bag of unrelated open pull requests across model additions, docs, tokenizer utilities, export tests, and loss/kernel fixes. The only true duplicate-looking pair is the two fixed_cross_entropy kwargs PRs targeting the same issue.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "#43251 is the clearest canonical PR: it targets the actual `fixed_cross_entropy` bug, has the explicit issue link (#43240), and includes the substantive API fix (`weight`/`label_smoothing` forwarding) plus test coverage.", + "canonical_pr_reason": "PR #43251 is the clearest canonical choice: it explicitly targets issue #43240, has the more complete fix surface, and includes review activity. PR #43254 is the near-duplicate follow-up variant.", "best_issue_reason": null, - "best_pr_reason": "#43251 is the best overall PR for the cluster because it is the only one that cleanly matches the shared underlying defect and is the most complete/maintainable fix among the duplicate pair.", + "best_pr_reason": "PR #43251 is the best representative of the cluster\u2019s only real duplicate pair because it directly addresses the tracked bug with the more complete implementation.", "soft_edge_verdicts": [ { "left": "pull_request:39895", "right": "pull_request:43424", "accept": false, - "reason": "Both are PRs, but they are about different features: VideoPrism model support vs ExecuTorch dynamic-shape export tests." + "reason": "Both are unrelated PRs; one adds VideoPrism, the other adds an Executorch dynamic-shape test." }, { "left": "pull_request:41855", - "right": "pull_request:43424", + "right": "pull_request:43823", "accept": false, - "reason": "Mistral tokenizer methods vs ExecuTorch export tests; no shared concrete code-path bug." + "reason": "Tokenizer helper methods for MistralCommon and a new MobileLLM model are different changes and code paths." }, { "left": "pull_request:41855", - "right": "pull_request:43823", + "right": "pull_request:43424", "accept": false, - "reason": "Mistral tokenizer API additions and MobileLLM model addition are unrelated changes." + "reason": "A tokenizer API addition and an Executorch export test are unrelated." }, { "left": "pull_request:42668", "right": "pull_request:43823", "accept": false, - "reason": "AudioFlamingo3 processor robustness vs MobileLLM model implementation; different code paths and goals." + "reason": "Processor auto-loading robustness and MobileLLM model implementation are not the same underlying change." }, { "left": "pull_request:43251", "right": "pull_request:43823", "accept": false, - "reason": "Loss-function kwargs forwarding vs adding MobileLLM; only superficial PR similarity." + "reason": "Cross-entropy kwargs plumbing and adding MobileLLM are unrelated." }, { "left": "pull_request:43251", "right": "pull_request:44827", "accept": false, - "reason": "Cross-entropy kwargs fix and Mistral4 test fixes are unrelated changes in different subsystems." + "reason": "The loss helper fix is unrelated to Mistral4 test/model adjustments." }, { "left": "pull_request:43424", "right": "pull_request:44827", "accept": false, - "reason": "ExecuTorch export test additions vs Mistral4 test fixes; no shared underlying bug." + "reason": "Executorch export tests and Mistral4 fixes do not overlap on the same bug." }, { "left": "pull_request:43424", "right": "pull_request:43823", "accept": false, - "reason": "ExecuTorch exportability tests and MobileLLM model addition are not the same change." + "reason": "Executorch dynamic-shape testing and MobileLLM model addition are separate concerns." }, { "left": "pull_request:43823", "right": "pull_request:45655", "accept": false, - "reason": "New MobileLLM model files vs fixing `cls.config` resolution order; unrelated code paths." + "reason": "A new model implementation and config resolution-order fix are different underlying changes." }, { "left": "pull_request:43823", "right": "pull_request:43995", "accept": false, - "reason": "MobileLLM implementation vs Falcon output-collection refactor; different model families and problems." + "reason": "MobileLLM model addition and Falcon output-interface refactor are unrelated." }, { "left": "pull_request:43995", "right": "pull_request:45213", "accept": false, - "reason": "Falcon refactor and model-creation skill are completely different changes." + "reason": "Falcon refactoring and the model-creation skill/docs PR are not the same code-path change." }, { "left": "pull_request:43995", "right": "pull_request:44827", "accept": false, - "reason": "Falcon standardized output refactor vs Mistral4 tests; not the same concrete bug or change." - }, - { - "left": "pull_request:44827", - "right": "pull_request:46495", - "accept": false, - "reason": "Mistral4 test fixes vs BLT beam-search/history-input device handling; different model-specific issues." + "reason": "Falcon output-capture refactor and Mistral4 test fixes are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:45170", "accept": false, - "reason": "Mistral4 tests and `layrnorm`->`layernorm` renaming are unrelated." + "reason": "Mistral4 test/model changes and the layrnorm->layernorm rename fix different problems." }, { "left": "pull_request:44827", "right": "pull_request:46267", "accept": false, - "reason": "Mistral4 test updates vs chat parsing/serve pipeline changes; no shared code-path problem." + "reason": "Mistral4 test fixes and chat parsing updates are unrelated." }, { "left": "pull_request:44827", "right": "pull_request:46048", "accept": false, - "reason": "Mistral4 tests vs TDT loss kernel integration; different functionality entirely." + "reason": "Mistral4 test fixes and the TDT loss kernel addition are different features." }, { "left": "pull_request:44827", - "right": "pull_request:45213", + "right": "pull_request:45655", "accept": false, - "reason": "Mistral4 tests and model-creation skill content are not duplicates." + "reason": "Mistral4 test changes and cls.config resolution-order fix are not the same bug." }, { - "left": "pull_request:43823", + "left": "pull_request:44827", "right": "pull_request:45213", "accept": false, - "reason": "MobileLLM model addition vs repo skill/documentation tooling; unrelated." + "reason": "A model test PR and a model-creation skill/docs PR are unrelated." }, { - "left": "pull_request:44827", - "right": "pull_request:45655", + "left": "pull_request:43823", + "right": "pull_request:45213", "accept": false, - "reason": "Mistral4 tests do not overlap with the `cls.config` resolution-order fix." + "reason": "MobileLLM model addition and the model-creation skill/docs PR are not duplicates." }, { "left": "pull_request:43251", "right": "pull_request:43254", "accept": true, - "reason": "These are the same underlying `fixed_cross_entropy` fix: both forward supported kwargs to `nn.functional.cross_entropy` for issue #43240, differing mainly in signature/style and completeness." + "reason": "Both PRs patch fixed_cross_entropy to forward cross_entropy kwargs for the same issue (#43240); they are duplicate implementations of the same fix." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: nearly all pairs are clearly unrelated, and the only accepted duplicate-like pair (#43251/#43254) is supported by the shared `src/transformers/loss/loss_utils.py` change and the shared issue target #43240. The canonical PR choice is conservative and justified by the more complete fix plus explicit issue link. No overstatement stands out." + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate pair, PR #43251/#43254, is supported by the shared issue target (#43240) and matching loss_utils fix. The other soft-edge rejections are consistent with the file/diff contents and do not overstate duplication." }, "evaluator_used": true, "retried": false, @@ -19312,7 +19180,7 @@ { "label": "pull_request", "component_index": 2, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19330,7 +19198,7 @@ "pull_request:40546|pull_request:42781", "pull_request:42781|pull_request:46331" ], - "prepared_review_unit_hash": "ecd09f1c6aafa1b1350ff6dc69e28de42f8161193b014e8b26969238e4fc7fb1", + "prepared_review_unit_hash": "82f4102e54dc3d6d535c096cc48e89a025aeb9728e1e4fb08ae42956f25e099e", "component_budget": { "node_count": 3, "item_count": 3, @@ -19352,24 +19220,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates of one another: PR 40546 adds the base VibeVoice model, PR 42781 adds a VibeVoice Realtime variant plus acoustic tokenizer, and PR 46331 adds Parakeet-RNNT. The shared files are mostly auto-registration/docs scaffolding rather than the same concrete code change.", - "confidence": 0.91, + "summary": "These PRs are not duplicates: one is the base VibeVoice implementation, one is a separate VibeVoice Realtime extension, and one is an unrelated Parakeet-RNNT change. The shared auto-registry files are broad integration overlap, not the same fix/change.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No issue is present; among the PRs, 40546 is the cleanest canonical representative because it is the core VibeVoice implementation, while 42781 is a broader realtime extension and 46331 is unrelated.", + "canonical_pr_reason": "PR 40546 is the core VibeVoice implementation and the most established candidate to represent the VibeVoice work; the other VibeVoice PR is a separate realtime extension, and the Parakeet PR is unrelated.", "best_issue_reason": null, - "best_pr_reason": "PR 40546 is the best representative of the cluster\u2019s main theme because it introduces the foundational VibeVoice model with a focused code footprint. PR 42781 expands into a separate realtime/acoustic-tokenizer feature set, and PR 46331 targets a different model family entirely.", + "best_pr_reason": "PR 40546 is the best representative of the cluster because it covers the main VibeVoice feature set and has the broadest, most mature implementation; the others are distinct follow-on or unrelated changes.", "soft_edge_verdicts": [ { "left": "pull_request:40546", "right": "pull_request:42781", "accept": false, - "reason": "Both are VibeVoice-related, but 40546 is the base model implementation while 42781 is a distinct VibeVoice Realtime feature set with additional acoustic-tokenizer work; they are not the same concrete change and would not plausibly merge into one PR." + "reason": "Related model family, but not the same underlying change: 40546 implements VibeVoice, while 42781 adds VibeVoice Realtime with additional acoustic-tokenizer and generation work. Shared registry/docs files are too broad to treat as duplicates." }, { "left": "pull_request:42781", "right": "pull_request:46331", "accept": false, - "reason": "These target different model families and different code paths. The overlap is limited to shared Transformers registration/docs plumbing, which is not enough to treat them as the same underlying bug or change." + "reason": "Different model targets and code paths: 42781 is VibeVoice Realtime, while 46331 adds Parakeet-RNNT. Overlap in auto-mapping files is incidental framework plumbing, not a duplicate fix." } ] }, @@ -19383,7 +19251,7 @@ { "label": "pull_request", "component_index": 3, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19399,12 +19267,12 @@ "soft_pairs": [ "pull_request:40755|pull_request:41224" ], - "prepared_review_unit_hash": "3c21b57b5eca2dfd8fc9f7bf6030e76132271a09748f509c50d1532399f63ca3", + "prepared_review_unit_hash": "278c89ecc870a6064c5f9f574556dd0da6db800d95ac84030b63849b941c2e20", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4676, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -19412,7 +19280,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4675, + "serialized_chars": 4676, "estimated_input_tokens": 1169, "estimated_eval_tokens": 2594 }, @@ -19421,18 +19289,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs are unrelated feature additions in different model families: TimesFM covariate forecasting vs DINOv3 image classification support. They do not describe the same underlying change and should not be clustered as duplicates.", - "confidence": 0.96, + "summary": "These two pull requests are not duplicates: one adds TimesFM covariate forecasting support, while the other adds DINOv3 image classification support. They touch different models, code paths, and user-facing features.", + "confidence": 0.99, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #41224 is the better canonical PR for this cluster because it is a concrete, self-contained feature addition with broader implementation changes and more review activity, making it the more representative item among the two unrelated PRs.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #41224 is the strongest representative of the set: it adds a specific DINOv3 image-classification capability with corresponding code and tests, whereas the TimesFM PR targets a separate forecasting feature in a different subsystem.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:40755", "right": "pull_request:41224", "accept": false, - "reason": "Different underlying changes in different models and code paths: TimesFM covariate forecasting vs DINOv3 image classification. They are not mergeable into one PR and are not duplicates." + "reason": "Different model families and unrelated changes: TimesFM covariate regression vs. DINOv3 image classification. They do not fix the same bug or implement the same feature." } ] }, @@ -19446,7 +19314,7 @@ { "label": "pull_request", "component_index": 4, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19509,7 +19377,7 @@ { "label": "pull_request", "component_index": 5, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19527,7 +19395,7 @@ "pull_request:41797|pull_request:45186", "pull_request:45186|pull_request:45630" ], - "prepared_review_unit_hash": "73619e00a5c69dafc5945448b0ea223d7cf0502022bb86ccc84d099dde29668e", + "prepared_review_unit_hash": "f06a1ded87d4ec8cef6c47b35a44ca6e0b18d293321460e6760131fb3fdbb510", "component_budget": { "node_count": 3, "item_count": 3, @@ -19549,24 +19417,24 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All three items are unrelated model-onboarding pull requests. They share Transformers boilerplate and auto-registry/docs files, but each adds a different model family (DeepSeek OCR, Isaac, Kimi2-6) with distinct codepaths and assets.", - "confidence": 0.91, + "summary": "These are three separate model-addition PRs for different models (DeepSeek OCR, Isaac, Kimi2-6). The overlap is mostly boilerplate around Transformers model registration and docs, not the same underlying change, so none should be clustered as duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45630 (Kimi2-6) is the strongest representative of the cluster: it has the broadest integration footprint and by far the most review activity, making it the best cluster anchor despite not being duplicate-related to the others.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR #45630 is the most complete and actively reviewed of the three model-addition PRs, so it's the best single representative for this cluster.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:41797", "right": "pull_request:45186", "accept": false, - "reason": "Both are generic new-model integrations, but they add different models with different configs, processors, and implementations. Shared files are only standard auto-registry/docs scaffolding, not the same underlying change." + "reason": "Different model integrations with different code paths and artifacts: DeepSeek OCR adds OCR-specific model/processor files, while Isaac adds a separate vision model stack and tests. Shared auto-registration/docs boilerplate is not enough to treat them as the same PR." }, { "left": "pull_request:45186", "right": "pull_request:45630", "accept": false, - "reason": "Isaac and Kimi2-6 are distinct model additions. The overlap is limited to common Transformers registration and documentation plumbing, so these are not the same code-path fix/change and should not be merged as duplicates." + "reason": "These add different models (Isaac vs Kimi2-6) with distinct configs, modeling code, and supporting files. The similarity is limited to generic model-registration updates, so they are not mergeable as one fix/change." } ] }, @@ -19580,7 +19448,7 @@ { "label": "pull_request", "component_index": 6, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19596,7 +19464,7 @@ "soft_pairs": [ "pull_request:42975|pull_request:42976" ], - "prepared_review_unit_hash": "02adde08c72e7fba72c8bc74fffb6e44b93ca82f69a58792436b74b18cded12e", + "prepared_review_unit_hash": "a0e2337b3dc977288bf39433d281c7b8fb32fddb91ae4ecf1f2a5321e3d9f34b", "component_budget": { "node_count": 2, "item_count": 2, @@ -19618,18 +19486,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related workflow-maintenance PRs, but they target different concrete upgrades: one is specifically for Node 24 compatibility across GitHub Actions, while the other broadly bumps action versions, including docker/build-push-action. They are not the same underlying change.", - "confidence": 0.9, + "summary": "These are two separate dependency-upgrade PRs in GitHub Actions workflows: one targets Node 24 compatibility for core actions, the other updates docker/build-push-action and related workflow actions. They overlap in files but not in the underlying change, so they should not be merged as duplicates.", + "confidence": 0.91, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 42975 is the more focused change set: it has a clear, specific goal (Node 24 compatibility) and a tighter scope across workflow action version bumps.", + "canonical_pr_reason": "No single canonical PR: #42975 and #42976 address different workflow-action upgrades with different scopes and failure modes.", "best_issue_reason": null, - "best_pr_reason": "PR 42975 is the best representative of the cluster because its intent is explicit and narrowly defined, making it the cleaner canonical artifact among the two.", + "best_pr_reason": "If one must be chosen, #42975 is the broader maintenance PR, but it still should not subsume #42976 because the actual action upgrades differ.", "soft_edge_verdicts": [ { "left": "pull_request:42975", "right": "pull_request:42976", "accept": false, - "reason": "Related action-version maintenance, but not the same bug/change: 42975 is about Node 24 compatibility, while 42976 upgrades different actions to latest versions, notably docker/build-push-action v6. They could coexist in separate PRs rather than be merged as duplicates." + "reason": "They both edit GitHub Actions workflows, but the concrete changes are different: #42975 upgrades checkout/cache/upload-artifact and other core actions for Node 24 compatibility, while #42976 upgrades docker/build-push-action and a different set of workflow actions. Shared files alone do not make them the same fix." } ] }, @@ -19643,7 +19511,7 @@ { "label": "pull_request", "component_index": 7, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19666,7 +19534,7 @@ "pull_request:43448|pull_request:45490", "pull_request:43451|pull_request:45490" ], - "prepared_review_unit_hash": "d4bda64b20444b7be1dd6a8dfe1de26dbfaf5665d299844901e789b06224b7e2", + "prepared_review_unit_hash": "f8ed0f1e01887f3627d90301ea393f1492c9bcfa420dac4b7e76dcf535fe2dc6", "component_budget": { "node_count": 4, "item_count": 4, @@ -19688,48 +19556,48 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "All four PRs are separate model-integration changes for different architectures (ViT NEPA, Molmo, Molmo2, CTSM). The overlap is only in shared boilerplate files like docs, auto-mappings, and package registration, so none of the soft pairs look like true duplicates.", - "confidence": 0.98, + "summary": "All four pull requests are separate model-addition efforts (ViT NEPA, Molmo, Molmo2, CTSM) with only superficial overlap in docs/auto-registration files. They do not look like duplicate fixes or mergeable into one PR.", + "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "No single PR is a clear duplicate-canonical representative; these are distinct model integrations rather than the same underlying change.", "soft_edge_verdicts": [ { "left": "pull_request:42978", "right": "pull_request:43448", "accept": false, - "reason": "Different model additions (ViT NEPA vs Molmo); shared docs/auto-registry files are generic integration boilerplate, not the same underlying change." + "reason": "Different model additions with different architectures and filenames; shared docs/auto-mapping edits are incidental, not the same code-path change." }, { "left": "pull_request:42978", "right": "pull_request:43451", "accept": false, - "reason": "ViT NEPA and Molmo2 are unrelated architectures with different codepaths; similarity is only from shared framework registration and docs." + "reason": "ViT NEPA vs Molmo2 are unrelated model integrations; overlap is only in generic registry/docs plumbing." }, { "left": "pull_request:42978", "right": "pull_request:45490", "accept": false, - "reason": "Completely different model families (ViT NEPA vs CTSM); no evidence they fix or add the same concrete code path." + "reason": "CTSM is a separate time-series model addition, not the same bug or feature as ViT NEPA." }, { "left": "pull_request:43448", "right": "pull_request:43451", "accept": false, - "reason": "Molmo and Molmo2 are distinct model families, and the shared files are expected new-model scaffolding rather than a duplicate change." + "reason": "Molmo and Molmo2 are related in naming but are distinct model families with different files and implementations; not a single concrete fix." }, { "left": "pull_request:43448", "right": "pull_request:45490", "accept": false, - "reason": "Molmo vs CTSM are unrelated model integrations; overlap in auto/model registration does not make them duplicates." + "reason": "Molmo and CTSM touch broad model-registration infrastructure, but they implement different model families and code paths." }, { "left": "pull_request:43451", "right": "pull_request:45490", "accept": false, - "reason": "Molmo2 and CTSM are separate model additions with different functionality; shared boilerplate edits are not enough to merge them." + "reason": "Molmo2 and CTSM are unrelated model additions; shared auto/model registry updates do not indicate duplicate content." } ] }, @@ -19743,7 +19611,7 @@ { "label": "pull_request", "component_index": 8, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19766,70 +19634,70 @@ "pull_request:43665|pull_request:46380", "pull_request:44815|pull_request:46380" ], - "prepared_review_unit_hash": "d5507072d9a5a5f478e5ced9d900b7140f063c452c85d9f53399cb094f1bd206", + "prepared_review_unit_hash": "6f6350727c5da31780616683841dc4cd98cc9170c4e053655eb1f8b6e3169e2c", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 11145, - "estimated_input_tokens": 2787, - "estimated_eval_tokens": 5830 + "serialized_chars": 11142, + "estimated_input_tokens": 2786, + "estimated_eval_tokens": 5828 }, "budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 6, - "serialized_chars": 11145, - "estimated_input_tokens": 2787, - "estimated_eval_tokens": 5830 + "serialized_chars": 11142, + "estimated_input_tokens": 2786, + "estimated_eval_tokens": 5828 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "No duplicate cluster here: the four PRs address different code paths and bug classes. 43291 is a Whisper tokenizer decode/test fix, 44815 is an FP8 dequant/loading fix, 46380 fixes watermarking config validation, and 43665 looks like unrelated test noise.", - "confidence": 0.91, + "summary": "No pair looks like the same underlying bug/change. The similarity is mostly superficial: one PR is test-only noise, and the others target unrelated areas (Whisper tokenization, FP8 dequantization, and generation watermarking validation).", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR for deduping; the PRs are materially unrelated and should remain separate.", + "canonical_pr_reason": "PR 46380 is the clearest real bug fix in the set: it addresses a specific validation/code-path issue by handling dict-form `watermarking_config` consistently in `GenerationConfig.validate()`.", "best_issue_reason": null, - "best_pr_reason": "PR 46380 is the strongest standalone fix: it patches a concrete runtime validation bug by converting dict watermarking configs before validation, with a clear code-path impact.", + "best_pr_reason": "PR 46380 is the strongest standalone candidate overall because it fixes a concrete runtime validation problem with a narrow, understandable change. PR 44815 is also substantive, but it is broader and more multi-part.", "soft_edge_verdicts": [ { "left": "pull_request:43291", "right": "pull_request:43665", "accept": false, - "reason": "Whisper tokenizer decode behavior vs. unrelated CLIP/ViT test edits; different subsystems and no shared concrete bug." + "reason": "Unrelated changes: 43291 modifies Whisper decode behavior, while 43665 is just test assertions/noise across CLIP and ViT." }, { "left": "pull_request:43291", "right": "pull_request:44815", "accept": false, - "reason": "Whisper decoding fix and FP8 dequant/loading fix are unrelated changes in different areas of the library." + "reason": "Different bug domains and code paths: Whisper tokenizer decoding vs FP8/dequant loading logic." }, { "left": "pull_request:43291", "right": "pull_request:46380", "accept": false, - "reason": "One is a Whisper tokenizer/test fix; the other is generation watermarking validation. Same repo, different bug." + "reason": "Whisper decode/tokenization changes are unrelated to generation watermarking validation." }, { "left": "pull_request:43665", "right": "pull_request:44815", "accept": false, - "reason": "Synthetic-looking test assertions in vision tests do not match the FP8 dequant/loading bug in 44815." + "reason": "43665 is test-only and not a real implementation fix; it does not match the FP8 dequantization change in 44815." }, { "left": "pull_request:43665", "right": "pull_request:46380", "accept": false, - "reason": "Unrelated test-only edits versus a generation config runtime fix; not the same underlying change." + "reason": "43665 is test-only noise, while 46380 fixes generation config validation." }, { "left": "pull_request:44815", "right": "pull_request:46380", "accept": false, - "reason": "Both are substantive, but they fix different concrete runtime problems in different subsystems and would not merge into one PR." + "reason": "Both are real fixes, but they target different problems and code paths: FP8 dequantization/loading vs watermarking config validation." } ] }, @@ -19843,7 +19711,7 @@ { "label": "pull_request", "component_index": 9, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19859,12 +19727,12 @@ "soft_pairs": [ "pull_request:43488|pull_request:43532" ], - "prepared_review_unit_hash": "36fd587943fcb228ca6a19b382472b08418b3a53de0ea3bc33fc84221b6a8f04", + "prepared_review_unit_hash": "d9df4361a0b23916649243c01601edca30858e13e4b3eb620d99cf84cd8558a2", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4562, + "serialized_chars": 4563, "estimated_input_tokens": 1141, "estimated_eval_tokens": 2538 }, @@ -19872,7 +19740,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4562, + "serialized_chars": 4563, "estimated_input_tokens": 1141, "estimated_eval_tokens": 2538 }, @@ -19881,18 +19749,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are unrelated pull requests: one is a small repo-bot/version/dependency test PR, the other changes CI/workflow and testing utilities. The shared Vit test file is incidental and not enough to treat them as the same change.", - "confidence": 0.98, + "summary": "Two open PRs that both touch the ViT test file, but they are otherwise unrelated: one is a small setup/version/dependency formatting change, the other rewires CI/workflow and testing utilities. The shared file alone is not enough to treat them as the same change.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #43532 is the more substantial and representative change in the cluster, but it is not a duplicate of #43488.", + "canonical_pr_reason": "PR #43532 is the more substantive and representative change in the cluster, with broader workflow/testing-utility edits beyond the shared test file.", "best_issue_reason": null, - "best_pr_reason": "PR #43532 is the better cluster representative because it has the broader, more substantive diff; however, it still addresses a different code path and should not be merged with #43488.", + "best_pr_reason": "PR #43532 is the best representative PR here because it has the larger, more distinctive code-path impact and clearer scope.", "soft_edge_verdicts": [ { "left": "pull_request:43488", "right": "pull_request:43532", "accept": false, - "reason": "Different purposes and code paths: #43488 is a setup/version/dependency test PR, while #43532 updates CI workflows and testing utilities. Shared Vit test file is incidental." + "reason": "Reject: they do not fix the same concrete bug/change. #43488 is a setup/version/dependency tweak, while #43532 changes CI workflow and testing utilities; the shared ViT test file is incidental." } ] }, @@ -19906,7 +19774,7 @@ { "label": "pull_request", "component_index": 10, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19922,12 +19790,12 @@ "soft_pairs": [ "pull_request:43612|pull_request:43613" ], - "prepared_review_unit_hash": "5e384244059950f2e4dd67d701563bcc738b05c6bda99f6c88a06a6461e51afc", + "prepared_review_unit_hash": "5ca9447d0d2dbebbd91dbbeb7883ea01927b314fe655b825e4cd4df6ec4a2f01", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6081, + "serialized_chars": 6083, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -19935,7 +19803,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6081, + "serialized_chars": 6083, "estimated_input_tokens": 1521, "estimated_eval_tokens": 3298 }, @@ -19944,18 +19812,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related but distinct pull requests for two different segmentation pipelines. They share scaffolding and documentation patterns, but they target different code paths, model integrations, and tests, so they should not be merged as duplicates.", - "confidence": 0.97, + "summary": "These two PRs share a lot of pipeline/doc scaffolding, but they implement different features: promptable concept segmentation vs promptable visual segmentation, with different model and test coverage. They are related, not duplicates.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "No single canonical PR for duplication: 43612 targets concept segmentation, while 43613 targets visual segmentation and touches a different set of models/code paths.", "best_issue_reason": null, - "best_pr_reason": "PR 43613 is the more substantial and broader implementation, with additional model/test coverage across SAM/EdgeTam/SAM2/SAM3 Tracker, so it is the better representative of this cluster even though it is not a duplicate of 43612.", + "best_pr_reason": "43613 is the better representative of the cluster because it is the broader, more complete PR and appears to define the more general visual segmentation pipeline.", "soft_edge_verdicts": [ { "left": "pull_request:43612", "right": "pull_request:43613", "accept": false, - "reason": "Different features: Promptable Concept Segmentation vs Promptable Visual Segmentation. Shared docs/registry scaffolding is expected, but the titles, task pages, pipeline modules, and model tests indicate separate concrete implementations rather than one bug or change." + "reason": "Both add similar pipeline/documentation scaffolding, but they implement different concrete features and model integrations, so they should not be merged as the same change." } ] }, @@ -19969,7 +19837,7 @@ { "label": "pull_request", "component_index": 11, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -19985,7 +19853,7 @@ "soft_pairs": [ "pull_request:43651|pull_request:43663" ], - "prepared_review_unit_hash": "ab717c2ed4b910d5f6faadd91c2f78c39197ea130a52b9d090dd3aeade54a70e", + "prepared_review_unit_hash": "e3ba5f58abafd35b8e20b1472489f7e11a535f1f9a5a68d1a16c7665f92e750e", "component_budget": { "node_count": 2, "item_count": 2, @@ -20007,18 +19875,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "The two PRs both touch `trainer.py`, but they add unrelated customization hooks: one changes gradient-accumulation loss scaling, the other refactors signature-column filtering for PEFT/custom trainers. They are not the same underlying fix.", + "summary": "These are two distinct trainer extensibility PRs in the same file: one changes gradient-accumulation loss scaling behavior, the other exposes signature-column filtering as an override point. They do not look mergeable into one PR or like duplicates.", "confidence": 0.97, "canonical_issue_reason": null, "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": null, + "best_pr_reason": "PR 43651 is the clearer standalone behavior change because it targets a concrete training-step loss-scaling path and introduces a single explicit control hook; 43663 is also valid, but it is a separate refactor/override point for data column filtering.", "soft_edge_verdicts": [ { "left": "pull_request:43651", "right": "pull_request:43663", "accept": false, - "reason": "Different trainer extension points and different bug classes: GA loss scaling vs. dataset column/signature filtering. They would not plausibly be merged as one PR." + "reason": "Same file, but different underlying behaviors: 43651 changes gradient accumulation loss normalization, while 43663 adds a hook for signature-based column filtering. They affect separate code paths and would not plausibly be merged as one fix." } ] }, @@ -20032,7 +19900,7 @@ { "label": "pull_request", "component_index": 12, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20048,7 +19916,7 @@ "soft_pairs": [ "pull_request:43656|pull_request:43842" ], - "prepared_review_unit_hash": "a53a41dc7f8537090dc4015080a2ba6ce3370c83170a92c66bd4fb840df70354", + "prepared_review_unit_hash": "48a14cf454e0a6403393ff46ed12435a2372962ac3b95088edb87ff0b3f0354f", "component_budget": { "node_count": 2, "item_count": 2, @@ -20070,18 +19938,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates: both touch `src/transformers/cli/serve.py` and both mention `TypeAdapter`, but they address different annotation/runtime-import problems with different fixes and different issue targets.", - "confidence": 0.84, + "summary": "These PRs touch the same CLI module but appear to fix different NameError/typing problems: 43656 is a broader annotation/future-import cleanup, while 43842 specifically guards a TypeAdapter reference so pydantic is not required at runtime. They are not the same underlying change.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 43842 is the narrower, more directly targeted fix for the `TypeAdapter` NameError when `pydantic` is absent, with a minimal `TYPE_CHECKING` import and quoted annotation.", + "canonical_pr_reason": "PR 43842 is the more precise representative because it targets a clearly scoped runtime NameError involving optional pydantic, whereas 43656 is a broader typing/annotation adjustment.", "best_issue_reason": null, - "best_pr_reason": "PR 43842 is the best representative of this cluster because it most clearly states and fixes the concrete runtime NameError scenario without broader annotation churn.", + "best_pr_reason": "PR 43842 is the strongest cluster representative: it has the clearest bug statement and the smallest, most targeted fix. 43656 looks like a separate typing-related change.", "soft_edge_verdicts": [ { "left": "pull_request:43656", "right": "pull_request:43842", "accept": false, - "reason": "Reject: same file and related annotation cleanup, but the diffs are different fixes for different reported issues/contexts; they do not clearly fix the same concrete bug and would not naturally merge into one PR." + "reason": "Same file and similar naming, but the fixes address different code-path problems: 43656 changes annotation behavior broadly, while 43842 specifically avoids a TypeAdapter runtime import/name error when pydantic is absent. They could not plausibly be merged as one duplicate PR." } ] }, @@ -20095,7 +19963,7 @@ { "label": "pull_request", "component_index": 13, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20161,7 +20029,7 @@ { "label": "pull_request", "component_index": 14, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 2, "status": "reviewed", @@ -20210,16 +20078,16 @@ "pull_request:44066|pull_request:44086", "pull_request:44066|pull_request:44071", "pull_request:44018|pull_request:44068", - "pull_request:44018|pull_request:44066", "pull_request:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", "pull_request:44019|pull_request:44722", "pull_request:44086|pull_request:44722", "pull_request:44019|pull_request:44085", "pull_request:44018|pull_request:44019", "pull_request:44019|pull_request:44071", "pull_request:44068|pull_request:44116", - "pull_request:44018|pull_request:44116", "pull_request:44068|pull_request:44722", + "pull_request:44018|pull_request:44116", "pull_request:44071|pull_request:44722", "pull_request:44018|pull_request:44085", "pull_request:44018|pull_request:44086", @@ -20227,21 +20095,21 @@ "pull_request:44071|pull_request:44085", "pull_request:44059|pull_request:44086", "pull_request:44085|pull_request:44086", - "pull_request:44098|pull_request:44101", "pull_request:44059|pull_request:44722", + "pull_request:44098|pull_request:44101", "pull_request:44018|pull_request:44071", - "pull_request:44017|pull_request:44722", "pull_request:44017|pull_request:44161", + "pull_request:44017|pull_request:44722", "pull_request:44017|pull_request:44073", "pull_request:44068|pull_request:44071", "pull_request:43997|pull_request:44073", "pull_request:44071|pull_request:44086", - "pull_request:44017|pull_request:44071", "pull_request:44073|pull_request:44101", + "pull_request:44017|pull_request:44071", "pull_request:44059|pull_request:44071", "pull_request:44073|pull_request:44161", - "pull_request:44018|pull_request:44059", "pull_request:44056|pull_request:44101", + "pull_request:44018|pull_request:44059", "pull_request:44059|pull_request:44068", "pull_request:44101|pull_request:44161", "pull_request:43997|pull_request:44017", @@ -20255,12 +20123,12 @@ "pull_request:44018|pull_request:44161", "pull_request:43997|pull_request:44071", "pull_request:44010|pull_request:44071", - "pull_request:44018|pull_request:44129", "pull_request:44010|pull_request:44073", + "pull_request:44018|pull_request:44129", "pull_request:43997|pull_request:44074", + "pull_request:43997|pull_request:44129", "pull_request:44071|pull_request:44129", "pull_request:44010|pull_request:44018", - "pull_request:43997|pull_request:44129", "pull_request:43996|pull_request:44161", "pull_request:43996|pull_request:44017", "pull_request:44010|pull_request:44044", @@ -20273,469 +20141,343 @@ "pull_request:44044|pull_request:44129", "pull_request:44044|pull_request:44154" ], - "prepared_review_unit_hash": "bc93da5e25beffe85ca1b83f6320cd9db3097ed1f6582a23297a8cbd54e5c440", + "prepared_review_unit_hash": "ed899cbd3b0969dea734fc8f4dd19142f25c553c68c0ab31386b888d04ee1bbc", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84896, - "estimated_input_tokens": 21224, - "estimated_eval_tokens": 42704 + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 }, "budget": { "node_count": 27, "item_count": 27, "soft_pair_count": 72, - "serialized_chars": 75091, - "estimated_input_tokens": 18773, - "estimated_eval_tokens": 37802 + "serialized_chars": 75095, + "estimated_input_tokens": 18774, + "estimated_eval_tokens": 37804 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "Mostly a loose cluster of model-specific output-tracing refactors that share the same pattern but not the same code path. The concrete duplicate pockets are GPT-J/CodeGen (44066/44722/44085) and GPT-Neo (44018/44068); the rest are separate architecture-specific changes.", - "confidence": 0.84, + "summary": "This cluster is mostly a theme cluster of per-model output-tracing refactors, not a single duplicate bug. The only strong duplicate-like matches are within the same model family/file (RegNet, GPT-Neo, GPT-J/CodeGen); cross-model pairs should be rejected.", + "confidence": 0.9, "canonical_issue_reason": null, - "canonical_pr_reason": "44066 is a solid representative for the GPT-J/CodeGen duplicate pocket because it directly spans the shared GPT-J and CodeGen files/code paths; I\u2019m not claiming it is broader than 44722, just that it cleanly anchors that overlap.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "44066 is the best overall PR anchor because it represents one of the two real duplicate pockets in the cluster (GPT-J/CodeGen) and has explicit issue linkage, while most other PRs are separate model-specific refactors.", + "best_pr_reason": "PR 44722 is the strongest representative of the GPT-J/CodeGen subset and the most complete late-stage refactor among the near-duplicates, but it does not represent the whole heterogeneous cluster.", "soft_edge_verdicts": [ { "left": "pull_request:43996", "right": "pull_request:44085", "accept": false, - "reason": "Different model families and forward paths; same tracking issue is not enough." + "reason": "Different model families; same umbrella output-tracing theme is not enough." }, { "left": "pull_request:43996", "right": "pull_request:44044", "accept": false, - "reason": "Different model implementations and separate output-tracing changes." + "reason": "CVT/FNet vs DeBERTa are unrelated model changes." }, { "left": "pull_request:44066", "right": "pull_request:44085", "accept": true, - "reason": "Same GPT-J output-tracing code path; these could plausibly be merged into one PR." + "reason": "Both are GPT-J output-tracing refactors in the same code path; 44066 is essentially a broader superset." }, { "left": "pull_request:44007", "right": "pull_request:44072", "accept": false, - "reason": "ResNet/RT-DETR versus EfficientNet are different architectures, not the same change." + "reason": "RegNet/ResNet family vs EfficientNet; different concrete changes." }, { "left": "pull_request:44072", "right": "pull_request:44722", "accept": false, - "reason": "EfficientNet and GPT-J/CodeGen are unrelated code paths." + "reason": "EfficientNet and GPT-J/CodeGen are unrelated." }, { "left": "pull_request:44066", "right": "pull_request:44072", "accept": false, - "reason": "GPT-J/CodeGen versus EfficientNet are different model-specific refactors." + "reason": "GPT-J/CodeGen vs EfficientNet are different model refactors." }, { "left": "pull_request:44013", "right": "pull_request:44044", "accept": false, - "reason": "MobileNetV2 and DeBERTa are unrelated model paths." + "reason": "MobileNetV2 and DeBERTa are unrelated." }, { "left": "pull_request:44066", "right": "pull_request:44086", "accept": false, - "reason": "GPT-J/CodeGen and MGP-STR are separate architectures and forward paths." + "reason": "GPT-J/CodeGen vs MGP-STR are unrelated changes." }, { "left": "pull_request:44066", "right": "pull_request:44071", "accept": false, - "reason": "GPT-J/CodeGen and MPT are different code paths." + "reason": "GPT-J/CodeGen vs MPT are unrelated." }, { "left": "pull_request:44018", "right": "pull_request:44068", "accept": true, - "reason": "Same GPT-Neo output-tracing refactor on the same file/code path." + "reason": "Same GPT-Neo output-tracing refactor in the same model/file; one PR is a superset-like variant." }, { - "left": "pull_request:44018", - "right": "pull_request:44066", + "left": "pull_request:44066", + "right": "pull_request:44068", "accept": false, - "reason": "GPT-Neo versus GPT-J/CodeGen are different model implementations." + "reason": "GPT-J/CodeGen and GPT-Neo are different model implementations." }, { - "left": "pull_request:44066", - "right": "pull_request:44068", + "left": "pull_request:44018", + "right": "pull_request:44066", "accept": false, - "reason": "GPT-J/CodeGen and GPT-Neo are different model code paths." + "reason": "GPT-Neo vs GPT-J/CodeGen are different concrete model paths." }, { "left": "pull_request:44019", "right": "pull_request:44722", "accept": false, - "reason": "ResNet and GPT-J/CodeGen are unrelated model families." + "reason": "ResNet and GPT-J/CodeGen are unrelated." }, { "left": "pull_request:44086", "right": "pull_request:44722", "accept": false, - "reason": "MGP-STR and GPT-J/CodeGen are separate concrete changes." + "reason": "MGP-STR and GPT-J/CodeGen are unrelated." }, { "left": "pull_request:44019", "right": "pull_request:44085", "accept": false, - "reason": "ResNet and GPT-J are different model paths." + "reason": "ResNet and GPT-J are unrelated." }, { "left": "pull_request:44018", "right": "pull_request:44019", "accept": false, - "reason": "GPT-Neo and ResNet do not share the same concrete code path." + "reason": "GPT-Neo and ResNet are unrelated model refactors." }, { "left": "pull_request:44019", "right": "pull_request:44071", "accept": false, - "reason": "ResNet and MPT are unrelated model implementations." + "reason": "ResNet and MPT are unrelated." }, { "left": "pull_request:44068", "right": "pull_request:44116", "accept": false, - "reason": "GPT-Neo and Flaubert/XLM are different model families." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44116", - "accept": false, - "reason": "GPT-Neo and Flaubert are separate output-tracing changes." + "reason": "GPT-Neo and Flaubert/XLM are unrelated." }, { "left": "pull_request:44068", "right": "pull_request:44722", "accept": false, - "reason": "GPT-Neo and GPT-J/CodeGen are different model code paths." - }, - { - "left": "pull_request:44071", - "right": "pull_request:44722", - "accept": false, - "reason": "MPT and GPT-J/CodeGen are unrelated implementations." + "reason": "GPT-Neo and GPT-J/CodeGen are different model families." }, { "left": "pull_request:44018", - "right": "pull_request:44085", - "accept": false, - "reason": "GPT-Neo and GPT-J are different model implementations." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44086", + "right": "pull_request:44059", "accept": false, - "reason": "GPT-Neo and MGP-STR are unrelated forward paths." + "reason": "GPT-Neo and GPT-2 are different concrete changes." }, { - "left": "pull_request:44071", + "left": "pull_request:44068", "right": "pull_request:44086", "accept": false, - "reason": "MPT and MGP-STR are different model-specific refactors." + "reason": "GPT-Neo and MGP-STR are unrelated." }, { "left": "pull_request:44059", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-2 and MGP-STR are unrelated model implementations." - }, - { - "left": "pull_request:44085", - "right": "pull_request:44086", - "accept": false, - "reason": "GPT-J and MGP-STR are different code paths." - }, - { - "left": "pull_request:44098", - "right": "pull_request:44101", + "right": "pull_request:44068", "accept": false, - "reason": "ViLT and Flaubert/XLM are separate model families." + "reason": "GPT-2 and GPT-Neo are separate model implementations." }, { "left": "pull_request:44059", - "right": "pull_request:44722", - "accept": false, - "reason": "GPT-2 and GPT-J/CodeGen are different model paths." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44722", + "right": "pull_request:44071", "accept": false, - "reason": "SegFormer and GPT-J/CodeGen are unrelated implementations." + "reason": "GPT-2 and MPT are unrelated." }, { - "left": "pull_request:44017", + "left": "pull_request:44073", "right": "pull_request:44161", "accept": false, - "reason": "SegFormer and LongT5 are different model paths." - }, - { - "left": "pull_request:44017", - "right": "pull_request:44073", - "accept": false, - "reason": "SegFormer and VisualBERT are unrelated concrete changes." + "reason": "VisualBert and LongT5 are unrelated." }, { - "left": "pull_request:44068", - "right": "pull_request:44071", + "left": "pull_request:43997", + "right": "pull_request:44017", "accept": false, - "reason": "GPT-Neo and MPT are different model implementations." + "reason": "RegNet and Segformer are different model code paths." }, { "left": "pull_request:43997", - "right": "pull_request:44073", + "right": "pull_request:44056", "accept": false, - "reason": "RegNet and VisualBERT are unrelated model paths." + "reason": "RegNet and MPNet are unrelated." }, { "left": "pull_request:44071", - "right": "pull_request:44076", + "right": "pull_request:44129", "accept": false, - "reason": "MPT and ImageGPT are different architectures." + "reason": "MPT and SpeechT5 are unrelated." }, { "left": "pull_request:44074", "right": "pull_request:44101", "accept": false, - "reason": "TextNet and Flaubert/XLM are different model families." + "reason": "TextNet and Flaubert/XLM are unrelated." }, { "left": "pull_request:44056", "right": "pull_request:44161", "accept": false, - "reason": "MPNet and LongT5 are separate output-handling changes." + "reason": "MPNet and LongT5 are unrelated." }, { "left": "pull_request:43997", "right": "pull_request:44018", "accept": false, - "reason": "RegNet and GPT-Neo are unrelated model implementations." + "reason": "RegNet and GPT-Neo are different model refactors." }, { "left": "pull_request:43997", "right": "pull_request:44161", "accept": false, - "reason": "RegNet and LongT5 are different model families." + "reason": "RegNet and LongT5 are unrelated." }, { - "left": "pull_request:44074", - "right": "pull_request:44161", + "left": "pull_request:43997", + "right": "pull_request:44071", "accept": false, - "reason": "TextNet and LongT5 are separate concrete changes." + "reason": "RegNet and MPT are unrelated." }, { "left": "pull_request:44010", "right": "pull_request:44071", "accept": false, - "reason": "SqueezeBERT and MPT are different model code paths." - }, - { - "left": "pull_request:44056", - "right": "pull_request:44101", - "accept": false, - "reason": "MPNet and Flaubert/XLM are unrelated implementations." + "reason": "SqueezeBert and MPT are unrelated." }, { "left": "pull_request:44010", "right": "pull_request:44073", "accept": false, - "reason": "SqueezeBERT and VisualBERT are different model families." - }, - { - "left": "pull_request:44044", - "right": "pull_request:44074", - "accept": false, - "reason": "DeBERTa and TextNet are unrelated output-tracing changes." - }, - { - "left": "pull_request:44066", - "right": "pull_request:44722", - "accept": true, - "reason": "Same GPT-J/CodeGen output-tracing refactor across the same concrete files and code paths." - }, - { - "left": "pull_request:44085", - "right": "pull_request:44722", - "accept": true, - "reason": "Same GPT-J output-tracing refactor in the same file; 44722 includes the same GPT-J path plus CodeGen." - }, - { - "left": "pull_request:44085", - "right": "pull_request:44129", - "accept": false, - "reason": "GPT-J and SpeechT5 are unrelated model code paths." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44007", - "accept": false, - "reason": "FNet/CVT and ResNet/RT-DETR are different model families." - }, - { - "left": "pull_request:43996", - "right": "pull_request:44129", - "accept": false, - "reason": "FNet/CVT and SpeechT5 are unrelated implementations." + "reason": "SqueezeBert and VisualBert are unrelated." }, { - "left": "pull_request:44044", + "left": "pull_request:44018", "right": "pull_request:44129", "accept": false, - "reason": "DeBERTa and SpeechT5 are different model paths." + "reason": "GPT-Neo and SpeechT5 are unrelated." }, { - "left": "pull_request:44013", + "left": "pull_request:43997", "right": "pull_request:44074", "accept": false, - "reason": "MobileNetV2 and TextNet are different architectures." - }, - { - "left": "pull_request:44073", - "right": "pull_request:44161", - "accept": false, - "reason": "VisualBERT and LongT5 are separate model-specific refactors." + "reason": "RegNet and TextNet are unrelated." }, { "left": "pull_request:43997", - "right": "pull_request:44017", + "right": "pull_request:44129", "accept": false, - "reason": "RegNet and SegFormer are different model families." + "reason": "RegNet and SpeechT5 are unrelated." }, { - "left": "pull_request:43997", - "right": "pull_request:44074", + "left": "pull_request:44071", + "right": "pull_request:44129", "accept": false, - "reason": "RegNet and TextNet are unrelated concrete changes." + "reason": "MPT and SpeechT5 are unrelated." }, { "left": "pull_request:44010", "right": "pull_request:44018", "accept": false, - "reason": "SqueezeBERT and GPT-Neo are different model paths." - }, - { - "left": "pull_request:44059", - "right": "pull_request:44068", - "accept": false, - "reason": "GPT-2 and GPT-Neo are unrelated model implementations." - }, - { - "left": "pull_request:44059", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-2 and MPT are different code paths." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44059", - "accept": false, - "reason": "GPT-Neo and GPT-2 are different model families." - }, - { - "left": "pull_request:44018", - "right": "pull_request:44071", - "accept": false, - "reason": "GPT-Neo and MPT are unrelated implementations." - }, - { - "left": "pull_request:44056", - "right": "pull_request:44071", - "accept": false, - "reason": "MPNet and MPT are different model code paths." + "reason": "SqueezeBert and GPT-Neo are unrelated." }, { - "left": "pull_request:43997", - "right": "pull_request:44056", + "left": "pull_request:43996", + "right": "pull_request:44161", "accept": false, - "reason": "RegNet and MPNet are separate model-specific refactors." + "reason": "CVT/FNet and LongT5 are unrelated." }, { - "left": "pull_request:44010", - "right": "pull_request:44018", + "left": "pull_request:43996", + "right": "pull_request:44017", "accept": false, - "reason": "SqueezeBERT and GPT-Neo are unrelated model paths." + "reason": "CVT/FNet and Segformer are unrelated." }, { "left": "pull_request:44010", "right": "pull_request:44044", "accept": false, - "reason": "SqueezeBERT and DeBERTa are different architectures." - }, - { - "left": "pull_request:43997", - "right": "pull_request:44056", - "accept": false, - "reason": "RegNet and MPNet are separate output-tracing changes." + "reason": "SqueezeBert and DeBERTa are unrelated." }, { - "left": "pull_request:44059", - "right": "pull_request:44086", + "left": "pull_request:44044", + "right": "pull_request:44074", "accept": false, - "reason": "GPT-2 and MGP-STR are unrelated model implementations." + "reason": "DeBERTa and TextNet are unrelated." }, { - "left": "pull_request:44098", - "right": "pull_request:44101", - "accept": false, - "reason": "ViLT and Flaubert/XLM are different concrete code paths." + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing refactor; 44722 is the broader later variant." }, { - "left": "pull_request:44073", - "right": "pull_request:44101", - "accept": false, - "reason": "VisualBERT and Flaubert/XLM are different model families." + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both touch the same GPT-J output-tracing code path; 44722 subsumes the narrower GPT-J-only variant." }, { - "left": "pull_request:44074", - "right": "pull_request:44101", + "left": "pull_request:44085", + "right": "pull_request:44129", "accept": false, - "reason": "TextNet and Flaubert/XLM are unrelated model paths." + "reason": "GPT-J and SpeechT5 are unrelated." }, { - "left": "pull_request:43997", - "right": "pull_request:44017", + "left": "pull_request:43996", + "right": "pull_request:44007", "accept": false, - "reason": "RegNet and SegFormer are different architectures." + "reason": "CVT/FNet vs ResNet/RegNet/RT-DETR are different model changes." }, { - "left": "pull_request:43997", - "right": "pull_request:44074", + "left": "pull_request:43996", + "right": "pull_request:44129", "accept": false, - "reason": "RegNet and TextNet are unrelated concrete changes." + "reason": "CVT/FNet and SpeechT5 are unrelated." }, { - "left": "pull_request:44010", - "right": "pull_request:44044", + "left": "pull_request:44044", + "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBERT and DeBERTa are different model families." + "reason": "DeBERTa and SpeechT5 are unrelated." }, { "left": "pull_request:44044", - "right": "pull_request:44074", + "right": "pull_request:44154", "accept": false, - "reason": "DeBERTa and TextNet are unrelated output-tracing refactors." + "reason": "DeBERTa and VITS are unrelated." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The soft-edge calls are mostly conservative, but the representative-pr/cluster rationale is not fully grounded. In the packet, PR 44066 is at least as broad as 44722 for the GPT-J/CodeGen overlap (both cover the same two files/code paths), so saying 44722 is the \"strongest\" or \"broadest\" representative is speculative. Also, the summary\u2019s phrasing that this is the \"only real overlap\" is a bit too strong given the separate GPT-Neo near-duplicate pair that is also present in the packet." + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats this as a heterogeneous output-tracing theme cluster rather than one bug, and the only accepted soft edges are the same-family/file overlaps that look mergeable (GPT-Neo and GPT-J/CodeGen supersets). Cross-model rejections are well supported by the diffs and titles." }, "evaluator_used": true, - "retried": true, + "retried": false, "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null @@ -20743,7 +20485,7 @@ { "label": "pull_request", "component_index": 14, - "component_count": 38, + "component_count": 35, "review_unit_index": 2, "review_unit_count": 2, "status": "reviewed", @@ -20770,14 +20512,14 @@ "pull_request:43997|pull_request:44010", "pull_request:44010|pull_request:44129" ], - "prepared_review_unit_hash": "3b6babea88d5d56b6ae2ae0521d9787e2a7fc097281fcdb416e8d33d1c852c2d", + "prepared_review_unit_hash": "a5365c581e2b86c4fe65d204e86afd9edc0c9322bc291242dd230a400aaa510a", "component_budget": { "node_count": 28, "item_count": 28, "soft_pair_count": 78, - "serialized_chars": 84896, - "estimated_input_tokens": 21224, - "estimated_eval_tokens": 42704 + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 }, "budget": { "node_count": 8, @@ -20792,48 +20534,48 @@ "aggressively_trimmed": false, "split": true, "analyst_result": { - "summary": "These PRs are all part of the same broad project to standardize output tracing/capture, but they target different model implementations and code paths, so they are not duplicates of one another.", - "confidence": 0.95, + "summary": "All items are separate model-specific refactors to standardized output tracing/capture; they share a theme but not the same underlying bug or change, so they should not be deduplicated together.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 44129 is the best canonical representative because it has the only explicit issue target and a fuller refactor description, but it is still a SpeechT5-specific implementation rather than a duplicate of the others.", + "canonical_pr_reason": "PR 44129 (SpeechT5) is the strongest standalone candidate because it has an explicit issue target, a concrete layer-level output-signature change, and a fuller implementation path than the other model migrations.", "best_issue_reason": null, - "best_pr_reason": "PR 44129 is the strongest single representative of the cluster: it clearly documents the standardized output-capture migration and is the only item tied to an explicit issue, making it the most informative anchor PR.", + "best_pr_reason": "PR 44129 is the best representative of the cluster as a concrete, issue-linked example of the output-capture refactor pattern, though it is not a duplicate of the others.", "soft_edge_verdicts": [ { "left": "pull_request:44024", "right": "pull_request:44076", "accept": false, - "reason": "Both are output-tracing refactors, but FocalNet and ImageGPT are different models with different forward paths and helper setup. Same theme, not the same concrete change." + "reason": "Both are output-tracing migrations, but they target different models (FocalNet vs ImageGPT) and different code paths; this is the same refactor pattern, not the same change." }, { "left": "pull_request:44073", "right": "pull_request:44074", "accept": false, - "reason": "VisualBERT and TextNet use different architectures and different output-capture rewrites. They are related refactors, but not one duplicate bugfix/change." + "reason": "Both standardize output collection, but one is VisualBert and the other is TextNet; they are unrelated model-specific edits, not one shared bug fix." }, { "left": "pull_request:44056", "right": "pull_request:44073", "accept": false, - "reason": "Both add standardized output capture, but MPNet and VisualBERT change different model internals and wrappers. They are not the same code-path fix." + "reason": "MPNet and VisualBert have similar decorator-based output tracing refactors, but they modify different architectures and hooks, so they are not the same underlying change." }, { "left": "pull_request:44056", "right": "pull_request:44074", "accept": false, - "reason": "MPNet and TextNet touch unrelated model implementations; the similarity is only at the refactor pattern level, not the underlying change." + "reason": "These are separate model migrations with different forward/output plumbing; similarity is only at the refactor pattern level." }, { "left": "pull_request:43997", "right": "pull_request:44010", "accept": false, - "reason": "RegNet and SqueezeBERT are distinct model code paths. They share the migration theme, but not a single underlying bug or mergeable change." + "reason": "RegNet and SqueezeBert changes both touch output recording infrastructure, but they do not fix the same concrete code path or bug." }, { "left": "pull_request:44010", "right": "pull_request:44129", "accept": false, - "reason": "SqueezeBERT and SpeechT5 are different models with different output/caching logic. This is the same refactor family, not a duplicate PR for one concrete fix." + "reason": "SqueezeBert and SpeechT5 are distinct model implementations with different output contracts; the overlap is only the standardized-output effort." } ] }, @@ -20847,7 +20589,7 @@ { "label": "pull_request", "component_index": 15, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -20885,13 +20627,17 @@ "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:44000|pull_request:44026", "pull_request:44001|pull_request:44002", - "pull_request:44001|pull_request:44003", "pull_request:44001|pull_request:44025", + "pull_request:44001|pull_request:44026", "pull_request:44002|pull_request:44029", + "pull_request:44002|pull_request:44025", + "pull_request:44001|pull_request:44003", "pull_request:44002|pull_request:44003", + "pull_request:44000|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", @@ -20912,304 +20658,328 @@ "pull_request:44028|pull_request:44029", "pull_request:44028|pull_request:44030", "pull_request:44029|pull_request:44030", - "pull_request:44000|pull_request:46438", "pull_request:44027|pull_request:46438", + "pull_request:44000|pull_request:46438", "pull_request:44029|pull_request:46438", "pull_request:44028|pull_request:46438", "pull_request:43998|pull_request:46438" ], - "prepared_review_unit_hash": "66d00a46691451b8aed6b6a52bd9231748dbf34d00cf3dea7a62a9a6368ca48d", + "prepared_review_unit_hash": "c0ee4af05efb8b2e8084b0884b7fc1eb6656ed4752fbd2e4bfae6f6b3fc3d1a8", "component_budget": { "node_count": 14, "item_count": 14, - "soft_pair_count": 44, - "serialized_chars": 43570, - "estimated_input_tokens": 10893, - "estimated_eval_tokens": 22042 + "soft_pair_count": 48, + "serialized_chars": 45107, + "estimated_input_tokens": 11277, + "estimated_eval_tokens": 22810 }, "budget": { "node_count": 14, "item_count": 14, - "soft_pair_count": 44, - "serialized_chars": 43570, - "estimated_input_tokens": 10893, - "estimated_eval_tokens": 22042 + "soft_pair_count": 48, + "serialized_chars": 45107, + "estimated_input_tokens": 11277, + "estimated_eval_tokens": 22810 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "A cluster of model-specific \"refactor output tracing\" PRs across different architectures. They share a pattern and utilities, but they do not appear to be duplicate fixes for the same concrete code path.", - "confidence": 0.95, + "summary": "This cluster is mostly a batch of similar-but-separate output-tracing refactors across different model implementations, plus one unrelated modular-conversion speedup PR. They share a pattern, but they are not the same concrete fix/change and should not be merged as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "pull_request:44003 is the most substantial representative in the cluster: it spans two related Mamba-family files and contains the clearest end-to-end output-tracing refactor with utility integration.", + "canonical_pr_reason": "No issue is present; as a representative PR, 44003 is the broadest of the refactor series because it updates two related Mamba-family model files and introduces the tracing/capture plumbing used by the rest of the set.", "best_issue_reason": null, - "best_pr_reason": "pull_request:44003 is the strongest anchor for this cluster because it is broader than the single-file edits and best represents the shared output-tracing refactor theme.", + "best_pr_reason": "PR 44003 is the best representative of the cluster\u2019s main theme: an output-tracing refactor in closely related model code. It is more central than the single-model PRs, while 46438 is unrelated.", "soft_edge_verdicts": [ { "left": "pull_request:43998", "right": "pull_request:44002", "accept": false, - "reason": "Different model families and forward paths (`timm_backbone` vs `upernet`); same refactor theme only." + "reason": "Different model families/files (timm_backbone vs upernet); same refactor pattern, not the same concrete change." }, { "left": "pull_request:43998", "right": "pull_request:44000", "accept": false, - "reason": "Different code paths (`timm_backbone` vs `vision_text_dual_encoder`); not the same bug/change." + "reason": "Different models and code paths; both refactor output handling, but not a single duplicate fix." }, { "left": "pull_request:43998", "right": "pull_request:44001", "accept": false, - "reason": "Unrelated model implementations; only a shared output-tracing refactor pattern." + "reason": "Different models/files; similar tracing cleanup, but not the same bug or mergeable PR." }, { "left": "pull_request:43998", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures (`timm_backbone` vs Mamba/Falcon-Mamba); not one mergeable fix." + "reason": "Different subsystems/models; one is timm_backbone, the other Mamba-family tracing changes." }, { "left": "pull_request:43998", "right": "pull_request:44029", "accept": false, - "reason": "Different model code paths (`timm_backbone` vs `rwkv`); superficial similarity only." + "reason": "Different architectures and files; only the high-level refactor theme matches." }, { "left": "pull_request:43999", "right": "pull_request:44001", "accept": false, - "reason": "Different models (`mobilenet_v1` vs `univnet`); not the same underlying change." + "reason": "Different models (MobileNetV1 vs UnivNet); shared output-tracing motif is too broad." }, { "left": "pull_request:43999", "right": "pull_request:44002", "accept": false, - "reason": "Different subsystems and file scopes; same refactor motif only." + "reason": "Different model code paths; not the same underlying fix." }, { "left": "pull_request:43999", "right": "pull_request:44025", "accept": false, - "reason": "`mobilenet_v1` and `depth_anything` are unrelated concrete code paths." + "reason": "Different models/files; both add can_return_tuple-style tracing, but they are separate changes." }, { "left": "pull_request:43999", "right": "pull_request:44026", "accept": false, - "reason": "Different wrapper/model code paths; cannot treat as one duplicate fix." + "reason": "Different encoder-decoder model wrappers; not one concrete duplicate change." }, { "left": "pull_request:43999", "right": "pull_request:44027", "accept": false, - "reason": "Different models and files; similar refactor, not the same change." + "reason": "Different model families; same broad refactor category only." }, { "left": "pull_request:44000", "right": "pull_request:44001", "accept": false, - "reason": "Different model families (`vision_text_dual_encoder` vs `univnet`); not mergeable as one fix." + "reason": "Vision-text dual encoder vs UnivNet; too different to be the same fix." }, { "left": "pull_request:44000", "right": "pull_request:44002", "accept": false, - "reason": "Different code paths and outputs; shared utility usage is not enough." + "reason": "Different model implementations; not mergeable as one PR." }, { "left": "pull_request:44000", - "right": "pull_request:44003", + "right": "pull_request:44025", "accept": false, - "reason": "Different architecture and implementation details; not the same concrete bug/change." + "reason": "Different tasks/models; both are output-tracing refactors, but not duplicates." }, { "left": "pull_request:44000", - "right": "pull_request:44025", + "right": "pull_request:44026", "accept": false, - "reason": "Different model families (`vision_text_dual_encoder` vs `depth_anything`)." + "reason": "Different model wrappers and files; same refactor pattern only." }, { "left": "pull_request:44001", "right": "pull_request:44002", "accept": false, - "reason": "Different models and code paths; only the refactor pattern matches." + "reason": "Unrelated model code paths; not the same concrete bug/change." }, { "left": "pull_request:44001", - "right": "pull_request:44003", + "right": "pull_request:44025", "accept": false, - "reason": "`univnet` and Mamba-family code are unrelated concrete paths." + "reason": "Different model families and files; too broad to treat as duplicate." }, { "left": "pull_request:44001", - "right": "pull_request:44025", + "right": "pull_request:44026", "accept": false, - "reason": "Different model implementations; not a single duplicate fix." + "reason": "Different model wrappers; same output-tracing theme, but separate PRs." }, { "left": "pull_request:44002", "right": "pull_request:44029", "accept": false, - "reason": "Different model families (`upernet` vs `rwkv`); same subsystem-level theme only." + "reason": "Different model architectures and files; not the same underlying code-path fix." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44025", + "accept": false, + "reason": "Different models; both refactor output handling, but not one shared change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and code paths; not a duplicate PR pair." }, { "left": "pull_request:44002", "right": "pull_request:44003", "accept": false, - "reason": "Different architectures and files; not one underlying bug or change." + "reason": "Different models/subsystems; only superficial similarity." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families; same general tracing motif is insufficient." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model code paths; both are refactors, not the same fix." }, { "left": "pull_request:44003", "right": "pull_request:44004", "accept": false, - "reason": "Different model families (`mamba`/`falcon_mamba` vs `codegen`); too broad to merge as one PR." + "reason": "Different model domains (Mamba vs CodeGen); not mergeable as one PR." }, { "left": "pull_request:43998", "right": "pull_request:44004", "accept": false, - "reason": "`timm_backbone` and `codegen` are unrelated code paths." + "reason": "Different models/files; only the output-tracing refactor pattern overlaps." }, { "left": "pull_request:44001", "right": "pull_request:44004", "accept": false, - "reason": "Different model implementations; only the refactor style is shared." + "reason": "Different model implementations; not the same concrete bug or change." }, { "left": "pull_request:44002", "right": "pull_request:44004", "accept": false, - "reason": "Different code paths and model logic; not the same change." + "reason": "Different models and code paths; too broad to dedupe." }, { "left": "pull_request:44000", "right": "pull_request:44004", "accept": false, - "reason": "Different architectures (`vision_text_dual_encoder` vs `codegen`)." + "reason": "Different architectures; same refactor theme but not the same PR." }, { "left": "pull_request:44025", "right": "pull_request:44026", "accept": false, - "reason": "Different encoder-decoder wrappers (`depth_anything` vs `vision_encoder_decoder`); not the same concrete fix." + "reason": "Related refactor style, but different model wrapper files and behavior." }, { "left": "pull_request:44025", "right": "pull_request:44027", "accept": false, - "reason": "Different model families (`depth_anything` vs `speech_encoder_decoder`)." + "reason": "Different model wrappers; not the same underlying change." }, { "left": "pull_request:44025", "right": "pull_request:44028", "accept": false, - "reason": "Different tasks and code paths (`depth_anything` vs `superpoint`)." + "reason": "Different tasks/models; output tracing cleanup alone is too broad." }, { "left": "pull_request:44025", "right": "pull_request:44029", "accept": false, - "reason": "Different model implementations; same output-tracing refactor theme only." + "reason": "Different model code paths; same pattern, not duplicate." }, { "left": "pull_request:44025", "right": "pull_request:44030", "accept": false, - "reason": "Different architectures (`depth_anything` vs `dpr`)." + "reason": "Different model families and files; not mergeable as one PR." }, { "left": "pull_request:44026", "right": "pull_request:44029", "accept": false, - "reason": "Different model families (`vision_encoder_decoder` vs `rwkv`)." + "reason": "Different seq2seq wrapper files; only the refactor motif matches." }, { "left": "pull_request:44026", "right": "pull_request:44027", "accept": false, - "reason": "Related wrapper shape, but different concrete code paths (`vision_encoder_decoder` vs `speech_encoder_decoder`)." + "reason": "Different model wrappers and code paths; not the same change." }, { "left": "pull_request:44026", "right": "pull_request:44028", "accept": false, - "reason": "Different model types (`vision_encoder_decoder` vs `superpoint`)." + "reason": "Different model domains; not a duplicate fix." }, { "left": "pull_request:44026", "right": "pull_request:44030", "accept": false, - "reason": "Different model families and forward contracts; not one duplicate fix." + "reason": "Different model code paths; shared tracing refactor is too generic." }, { "left": "pull_request:44027", "right": "pull_request:44028", "accept": false, - "reason": "Different model implementations and output contracts." + "reason": "Different model families (speech encoder-decoder vs SuperPoint); not the same concrete bug/change." }, { "left": "pull_request:44027", "right": "pull_request:44029", "accept": false, - "reason": "Different architectures (`speech_encoder_decoder` vs `rwkv`)." + "reason": "Different model implementations; not mergeable as one PR." }, { "left": "pull_request:44027", "right": "pull_request:44030", "accept": false, - "reason": "Different code paths; similar refactor pattern only." + "reason": "Different architectures; same high-level refactor theme only." }, { "left": "pull_request:44028", "right": "pull_request:44029", "accept": false, - "reason": "Different model families (`superpoint` vs `rwkv`)." + "reason": "Different model code paths; not a duplicate." }, { "left": "pull_request:44028", "right": "pull_request:44030", "accept": false, - "reason": "Different tasks and files; not the same underlying change." + "reason": "Different model families; output-tracing cleanup is not enough to combine them." }, { "left": "pull_request:44029", "right": "pull_request:44030", "accept": false, - "reason": "Different model implementations (`rwkv` vs `dpr`); broad similarity only." + "reason": "Different models/subsystems; similar refactor category, but separate changes." }, { - "left": "pull_request:44000", + "left": "pull_request:44027", "right": "pull_request:46438", "accept": false, - "reason": "Model refactor PR vs tooling/performance PR; unrelated changes." + "reason": "Unrelated changes: speech encoder-decoder output tracing vs modular-conversion process-pool reuse." }, { - "left": "pull_request:44027", + "left": "pull_request:44000", "right": "pull_request:46438", "accept": false, - "reason": "Different domains entirely (`speech_encoder_decoder` vs modular conversion tooling)." + "reason": "Unrelated model refactor vs build/tooling performance change." }, { "left": "pull_request:44029", "right": "pull_request:46438", "accept": false, - "reason": "`rwkv` model code and build tooling are unrelated." + "reason": "Completely different code paths; one is RWKV tracing, the other a modular conversion utility optimization." }, { "left": "pull_request:44028", "right": "pull_request:46438", "accept": false, - "reason": "Different subsystem and purpose; not the same bug/change." + "reason": "Different domains and purposes; not the same underlying change." }, { "left": "pull_request:43998", "right": "pull_request:46438", "accept": false, - "reason": "Model output-tracing refactor vs tooling pool reuse; unrelated." + "reason": "Model refactor and tooling optimization are unrelated." } ] }, @@ -21223,7 +20993,7 @@ { "label": "pull_request", "component_index": 16, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21239,12 +21009,12 @@ "soft_pairs": [ "pull_request:44662|pull_request:45401" ], - "prepared_review_unit_hash": "2c56715a847b277072db8e870b0017d90137d360dbb0ab38228c63c1c3b75e05", + "prepared_review_unit_hash": "84c3470a75dea818b4949d854d0fb39aa539da9bcbc3c95b489e1c3b842fd657", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6549, + "serialized_chars": 6551, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -21252,7 +21022,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 6549, + "serialized_chars": 6551, "estimated_input_tokens": 1638, "estimated_eval_tokens": 3532 }, @@ -21261,18 +21031,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two unrelated model-addition PRs that only overlap on standard Transformers scaffolding (auto registries, docs, package init). They do not appear to fix the same bug or implement the same change.", - "confidence": 0.97, + "summary": "These are two unrelated model-addition PRs: one implements PenguinVL, the other adds Voxtral TTS. The overlap is limited to standard auto-registration/docs scaffolding, not the same underlying change.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "If a PR must represent the cluster, #45401 is the better candidate because it is scoped to a single Voxtral-TTS feature and explicitly targets issue #45027; #44662 is a separate PenguinVL implementation.", + "canonical_pr_reason": "No single canonical PR: the changes target different models, different code paths, and would not be merged as one patch.", "best_issue_reason": null, - "best_pr_reason": "#45401 is the strongest standalone PR: it has an explicit issue target, a focused feature goal, and a coherent set of Voxtral-TTS files. #44662 is a different model implementation and not a duplicate.", + "best_pr_reason": "PR 45401 is the cleaner representative if one must be picked because it has an explicit issue target and a narrower, self-contained Voxtral TTS scope.", "soft_edge_verdicts": [ { "left": "pull_request:44662", "right": "pull_request:45401", "accept": false, - "reason": "Reject: PenguinVL and Voxtral TTS are different model implementations with different code paths and different added files. The shared registry/docs changes are boilerplate, not evidence of the same underlying change." + "reason": "Different model integrations (PenguinVL vs Voxtral TTS); shared files are generic registration/docs plumbing, not the same bug or feature." } ] }, @@ -21286,7 +21056,7 @@ { "label": "pull_request", "component_index": 17, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21306,12 +21076,12 @@ "pull_request:45363|pull_request:46339", "pull_request:45666|pull_request:46339" ], - "prepared_review_unit_hash": "17a92fd0bd4efd2a166a4303dcab3ec71f133bde76f05fda80ebaae0db76de30", + "prepared_review_unit_hash": "bc9a53cba695ef23ec5e78cb64843558cd22457fb2effde61f9db1ff1f25677e", "component_budget": { "node_count": 4, "item_count": 4, "soft_pair_count": 3, - "serialized_chars": 9634, + "serialized_chars": 9633, "estimated_input_tokens": 2409, "estimated_eval_tokens": 5074 }, @@ -21319,7 +21089,7 @@ "node_count": 4, "item_count": 4, "soft_pair_count": 3, - "serialized_chars": 9634, + "serialized_chars": 9633, "estimated_input_tokens": 2409, "estimated_eval_tokens": 5074 }, @@ -21328,47 +21098,47 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These PRs cluster around the same kernel-fusion area, but they split into a foundational Module Fusion API PR and several successive KernelConfig-based implementation/refinement PRs. #46339 is still the best representative of the later n-to-1 KernelConfig line, but the pairwise soft edges are not strong enough to treat the PRs as the same concrete change.", - "confidence": 0.74, + "summary": "All four items are closely related kernel-fusion PRs, but they are not all the same change. The clearest canonical PR is the latest, most complete n-to-1 fusion refinement (46339).", + "confidence": 0.79, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46339 is the best representative of the KernelConfig n-to-1 fusion line because it is the latest and broadest of the related implementation PRs, touching the core loading/integration path and tests, so it most likely subsumes the earlier KernelConfig iterations.", + "canonical_pr_reason": "PR 46339 is the most recent and most comprehensive refinement of the n-to-1 KernelConfig fusion work, with broader edits in the same loading/integration path and tests, so it is the best representative of the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR #46339 is the strongest cluster representative for the later kernel-fusion work: it is the most complete KernelConfig-based PR and appears to be the endpoint of the refinement chain, unlike #44979 which is a separate foundational API layer.", + "best_pr_reason": "PR 46339 best captures the cluster\u2019s main change set: it extends and simplifies the same KernelConfig-based n-to-1 fusion mechanism across core loading, hub kernels, modeling utils, and tests.", "soft_edge_verdicts": [ { "left": "pull_request:44979", "right": "pull_request:45363", "accept": false, - "reason": "Same general feature family, but #44979 is a foundational Module Fusion API PR while #45363 is a specific KernelConfig n-to-1 implementation. They are related layers, not the same concrete change." + "reason": "Related feature area, but 44979 introduces a new Module Fusion API while 45363 implements n-to-1 KernelConfig fusion; these are adjacent foundations, not the same concrete PR-level change." }, { "left": "pull_request:45363", "right": "pull_request:46339", - "accept": false, - "reason": "They overlap in KernelConfig fusion files and titles, but the evidence looks like successive iterations/extensions of the feature, not a clearly mergeable duplicate bugfix or identical PR." + "accept": true, + "reason": "Both target the same concrete n-to-1 KernelConfig fusion path in the model-loading/integration stack, and 46339 reads like a direct refinement of the same feature work." }, { "left": "pull_request:45666", "right": "pull_request:46339", - "accept": false, - "reason": "#45666 and #46339 are both KernelConfig fusion PRs, but the overlap is consistent with refinement of the same feature area rather than the same exact change set or duplicate fix." + "accept": true, + "reason": "These appear to be successive revisions of the same n-to-1 KernelConfig fusion implementation, touching the same code paths and sharing the same underlying feature goal." } ] }, "evaluator_result": { - "accept": false, - "feedback": "The high-level grouping is mostly grounded: #44979 is a separate foundational Module Fusion API PR, while #45363/#45666/#46339 are all in the KernelConfig n-to-1 fusion line. However, the soft-edge accepts are a bit too aggressive for the packet. The evidence only shows title/file overlap and later iterations, not that #45363\u2194#46339 or #45666\u2194#46339 are clearly mergeable as the same bugfix/duplicate PR. They look more like successive extensions/refinements of the same feature area than the same concrete change, so those pair verdicts should be more conservative." + "accept": true, + "feedback": "The summary is grounded in the packet: 46339 is the latest and broadest n-to-1 KernelConfig fusion refinement, and the soft-edge verdicts stay conservative. The 44979 vs 45363 rejection is appropriately cautious, while the 45363\u219446339 and 45666\u219446339 accepts are supported by overlapping filenames and the shared feature path." }, "evaluator_used": true, - "retried": true, - "accepted_nontrivial_soft_edge": false, + "retried": false, + "accepted_nontrivial_soft_edge": true, "error_kind": null, "error_message": null }, { "label": "pull_request", "component_index": 18, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21384,7 +21154,7 @@ "soft_pairs": [ "pull_request:45086|pull_request:45317" ], - "prepared_review_unit_hash": "17fc34ec7e9462093e02dd1f2ef3ce4bbb9b15ead4bad0e95cf1b9cbbf1b13e1", + "prepared_review_unit_hash": "335c8e09379e95a12adb80a8a99579ade430794451ec62563bb49c1380f6064b", "component_budget": { "node_count": 2, "item_count": 2, @@ -21406,18 +21176,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs implement the same fix for the Mistral regex patch AttributeError in the same file and on the same issue; 45317 is the more complete version because it includes the test coverage as well.", - "confidence": 0.97, + "summary": "Both PRs fix the same AttributeError in `_patch_mistral_regex` by changing the pre-tokenizer access from `backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; 45317 is the more complete version because it adds a test and appears to be the later refinement of the same fix.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45317 is the stronger canonical PR: it targets the same bug as 45086, but it also adds tests and has the clearer, more specific title tied to the fix.", + "canonical_pr_reason": "PR 45317 is the better canonical PR because it targets the same bug as 45086 but includes a test change and looks like the more complete, later fix.", "best_issue_reason": null, - "best_pr_reason": "PR 45317 is the best representative of the cluster because it contains the same code-path fix plus accompanying test coverage, making it the most complete implementation.", + "best_pr_reason": "PR 45317 is the best representative because it directly fixes the AttributeError and adds coverage, making it the clearest and most complete implementation of the change.", "soft_edge_verdicts": [ { "left": "pull_request:45086", "right": "pull_request:45317", "accept": true, - "reason": "Same underlying AttributeError fix in src/transformers/tokenization_utils_tokenizers.py for _patch_mistral_regex; the second PR is an expanded version of the first rather than a different change." + "reason": "Same concrete bug fix in the same code path and same issue target; 45317 is essentially the follow-up/expanded version of 45086." } ] }, @@ -21431,7 +21201,7 @@ { "label": "pull_request", "component_index": 19, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21447,7 +21217,7 @@ "soft_pairs": [ "pull_request:45115|pull_request:45569" ], - "prepared_review_unit_hash": "2e4f68141471b9f117c5c981f44ba629486d2e2be29e3fcc75c872e0bf881f37", + "prepared_review_unit_hash": "59007ab6f309b9a856c6f05ff28270fdad6ddc0c19e91dc5512750923e3f1247", "component_budget": { "node_count": 2, "item_count": 2, @@ -21469,18 +21239,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests are about Nemotron-H, but they are not the same change: one is a narrow inheritance/conversion-mapping refactor, while the other introduces the dedicated dense/sparse model split plus docs and auto-mapping updates. I would not merge them as duplicates.", - "confidence": 0.92, + "summary": "These PRs are related NemotronH work, but they target different end states: one refactors the existing NemotronH implementation to inherit GraniteMoeHybrid-style conversion/mapping, while the other splits NemotronH into dedicated dense/sparse variants and turns nemotron_h into a compatibility dispatcher. They overlap in shared model files, but not in the same concrete fix.", + "confidence": 0.82, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45569 is the better canonical representative because it is the larger, more complete change set: it introduces the new NemotronH dense/sparse model structure and updates the surrounding docs/auto-registration, whereas 45115 is a narrower refactor focused on inheritance and conversion naming.", + "canonical_pr_reason": "PR #45569 is the better cluster representative because it is the broader, newer architectural change that introduces the dense/sparse split and compatibility layer, whereas #45115 is a narrower refactor/conversion mapping update.", "best_issue_reason": null, - "best_pr_reason": "PR 45569 best captures the underlying user-visible change and is the more complete implementation of the Nemotron-H model split; 45115 looks like a preparatory refactor that overlaps only in shared model files.", + "best_pr_reason": "PR #45569 is the strongest standalone PR to keep: it contains the more complete user-facing model restructuring and associated docs/tests, making it the most meaningful canonical artifact for NemotronH changes.", "soft_edge_verdicts": [ { "left": "pull_request:45115", "right": "pull_request:45569", "accept": false, - "reason": "Overlap is only in the shared Nemotron-H codebase and auto/modeling files. 45115 is a conversion/inheritance refactor, while 45569 adds the dense/sparse model variants and documentation; they are not the same concrete fix and would not plausibly be merged as one PR." + "reason": "They overlap on NemotronH internals, but #45115 is a refactor/conversion-mapping change and #45569 is a model-splitting/dispatcher change. They are related work, not the same concrete bug or mergeable duplicate PR." } ] }, @@ -21494,7 +21264,7 @@ { "label": "pull_request", "component_index": 20, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21510,7 +21280,7 @@ "soft_pairs": [ "pull_request:45332|pull_request:45333" ], - "prepared_review_unit_hash": "8d9583610fd9c802c0499026637edde43633857e819a19ee2f47f164a474caf9", + "prepared_review_unit_hash": "f5e7416d7752738b0b1ea74cb13cb2dca999f49a3d95fbfc2645f0c111be5e23", "component_budget": { "node_count": 2, "item_count": 2, @@ -21532,18 +21302,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related but distinct feature PRs: one adds heterogeneous per-layer config support, the other builds broader heterogeneous modeling/runtime support on top of it. They overlap in infrastructure files, but they are not the same concrete change and should not be merged as duplicates.", - "confidence": 0.91, + "summary": "These are related feature PRs for heterogeneous model support, but they are not duplicates: one adds per-layer configuration, while the other adds downstream modeling/runtime support that consumes it.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45332 is the better canonical PR because it is the broader end-to-end implementation, covering per-layer config handling plus cache/modeling/runtime integration, whereas 45333 is just the configuration foundation.", + "canonical_pr_reason": "PR 45332 is the broader end-to-end change: it covers heterogeneous modeling/runtime behavior across cache, masking, modeling, and integrations, so it is the better representative of the cluster.", "best_issue_reason": null, - "best_pr_reason": "PR 45332 is the strongest representative of the cluster because it subsumes the shared heterogeneous-config work and extends it into actual model support and tests.", + "best_pr_reason": "PR 45332 best captures the underlying feature work because it includes the concrete model-path changes that make heterogeneous configs actionable; 45333 is a narrower foundational config-only piece.", "soft_edge_verdicts": [ { "left": "pull_request:45332", "right": "pull_request:45333", "accept": false, - "reason": "Related feature work, but not the same underlying change: 45333 focuses on per-layer configuration plumbing, while 45332 adds broader heterogeneous model/runtime support and can stand as a separate PR." + "reason": "Related but not the same concrete change: 45333 adds per-layer config support, while 45332 adds model/runtime support that uses heterogeneous config. They could be part of the same feature area, but they are not a duplicate fix and would not naturally be merged as one PR without broadening scope." } ] }, @@ -21557,7 +21327,7 @@ { "label": "pull_request", "component_index": 21, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21573,7 +21343,7 @@ "soft_pairs": [ "pull_request:45415|pull_request:45425" ], - "prepared_review_unit_hash": "0a142b77303974f8287d535ebdc509c5e291ebaa1d2e832c1464982aa189aa6b", + "prepared_review_unit_hash": "13b486fabb43328c791425d13f652cd95991b6a8f310b881863ccc113a414300", "component_budget": { "node_count": 2, "item_count": 2, @@ -21595,18 +21365,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are typing-focused PRs in transformers, but they are not the same change: one is a broad repo-wide type-checking sweep, the other is a narrower typing update around modeling_utils/_typing. The overlap in files and theme is not enough to treat them as duplicates.", - "confidence": 0.92, + "summary": "Both PRs are typing-related, but they are not the same change: #45415 is a broad type-checking sweep across many `src/transformers/*` files plus `pyproject.toml`, while #45425 adds a few concrete typing protocols/aliases in `_typing.py` and updates a small set of consumers. The overlap is thematic, not a duplicate code-path fix.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": null, + "canonical_pr_reason": "#45425 is the more concrete, self-contained typing change; #45415 is a much broader repository-wide type-checking cleanup rather than the same underlying fix.", "best_issue_reason": null, - "best_pr_reason": "PR #45425 is the more focused and concrete typing change, centered on _typing/modeling_utils integration, whereas #45415 is a much broader type-checking sweep across many files.", + "best_pr_reason": "#45425 best represents the cluster because it introduces specific typing abstractions and targeted call-site updates, whereas #45415 is a broad static-analysis sweep with many unrelated file edits.", "soft_edge_verdicts": [ { "left": "pull_request:45415", "right": "pull_request:45425", "accept": false, - "reason": "They share typing-related files, but the scope differs: #45415 is a wide type-checking pass over many modules, while #45425 is a targeted typing adjustment. They do not look like the same underlying fix." + "reason": "Shared typing infrastructure and a few filenames do not make these the same PR: one is a wide pyright/type-ignore sweep, the other is a focused protocol/alias addition for model utilities and PEFT integration." } ] }, @@ -21620,7 +21390,7 @@ { "label": "pull_request", "component_index": 22, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21636,12 +21406,12 @@ "soft_pairs": [ "pull_request:45513|pull_request:46084" ], - "prepared_review_unit_hash": "ad14178b58e50dc3e9ea6ffdfe617eac5fa52757b8f5687b4995ceb6fb92ce74", + "prepared_review_unit_hash": "b37d389367a3ec7581d2b4a2ecd975d54696e7949b19b279b055793904beb97f", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4857, + "serialized_chars": 4858, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -21649,7 +21419,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4857, + "serialized_chars": 4858, "estimated_input_tokens": 1215, "estimated_eval_tokens": 2686 }, @@ -21658,18 +21428,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both pull requests are about cache-aware multi-token forward paths, but they target different model families and different concrete bugs, so they should not be merged as duplicates.", - "confidence": 0.91, + "summary": "These are two unrelated open pull requests fixing different cache-related forward-pass bugs in different model families: Qwen3.5/Qwen3.5-MoE/Qwen3-Next linear attention vs Mamba2 CUDA-kernel cache handling.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #45513 is the better cluster representative: it addresses a broader cache-forward bug across several related Qwen/Olmo model implementations and includes model/test coverage for the multi-token cached continuation path.", + "canonical_pr_reason": "PR #46084 is the clearest representative: it has an explicit issue target, a focused fix for a specific seq_len>1 cache path in Mamba2, and a tight file set.", "best_issue_reason": null, - "best_pr_reason": "PR #45513 is the strongest overall example of the cluster\u2019s theme because it fixes a concrete cached-forward edge case in a more general way across multiple files, whereas #46084 is a narrower Mamba2-specific cache-length fix.", + "best_pr_reason": "PR #46084 is the best standalone representative because it is narrowly scoped, explicitly tied to issue #46032, and fixes one concrete code path. PR #45513 is broader and spans several model variants, but it addresses a different bug.", "soft_edge_verdicts": [ { "left": "pull_request:45513", "right": "pull_request:46084", "accept": false, - "reason": "They are related only at a high level (cache/use_cache behavior) but fix different model-specific code paths: Qwen/Olmo linear attention multi-token cached forward vs. Mamba2 seq_len>1 cached state handling. Not the same underlying bug or mergeable as one PR." + "reason": "They touch different models and different cache bugs. #45513 fixes multi-token cached forward in Qwen3.5/Qwen3-Next/Olmo Hybrid linear attention, while #46084 fixes Mamba2 CUDA-kernel use_cache behavior when seq_len > 1. They are not the same underlying change and would not plausibly merge into one PR." } ] }, @@ -21683,7 +21453,7 @@ { "label": "pull_request", "component_index": 23, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21699,7 +21469,7 @@ "soft_pairs": [ "pull_request:45548|pull_request:45662" ], - "prepared_review_unit_hash": "cadee3a3ecef72ba2d8df360182e8e59eb344cae81e52c7d2ef6dedeb779ccb0", + "prepared_review_unit_hash": "84dc521abaf91e273548ca86d8610fdb89da22c647216d9e5d52e0afcbf2d946", "component_budget": { "node_count": 2, "item_count": 2, @@ -21721,18 +21491,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs touch expert-parallel loading, but they fix different bugs: 45548 preserves distributed_config/EP detection for accelerate+ZeRO-3 loading, while 45662 prevents expert tensors from being overwritten in the FSDP2/sharded load path.", - "confidence": 0.96, + "summary": "The two PRs are thematically related (expert parallelism loading), but they fix different bugs: one preserves/recognizes distributed_config for EP+DeepSpeed ZeRO-3 loading, while the other prevents expert parameters from being overwritten during EP+FSDP2 sharded loading. They should not be treated as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "If a single PR must anchor this cluster, 45662 is the better representative because it fixes a concrete silent corruption bug in the loading path; 45548 is a separate config/EP-detection change.", + "canonical_pr_reason": "PR 45662 is the more concrete representative of the cluster because it fixes a specific loading overwrite path in sharded/distributed parameter handling, with a clear code-path bug and targeted remediation.", "best_issue_reason": null, - "best_pr_reason": "45662 is the stronger standalone PR: it targets a specific expert-overwrite failure mode with direct code-path fixes, whereas 45548 is narrower and addresses a different loading/config serialization problem.", + "best_pr_reason": "PR 45662 is the strongest standalone PR because it addresses a clearly defined failure mode (expert parameters being silently overwritten during post-shard loading) and is less of a supporting/configuration fix than 45548.", "soft_edge_verdicts": [ { "left": "pull_request:45548", "right": "pull_request:45662", "accept": false, - "reason": "Different root causes and fixes: 45548 handles EP config serialization/has_ep for ZeRO-3 loading; 45662 handles post-shard wrapping to stop experts being overwritten under FSDP2. Shared subsystem only, not the same change." + "reason": "Both concern EP loading, but 45548 is about serializing/discovering distributed_config for accelerate/ZeRO-3, while 45662 fixes a different overwrite bug in post-shard parameter assignment under FSDP2. Different concrete code paths, so not a duplicate." } ] }, @@ -21746,7 +21516,7 @@ { "label": "pull_request", "component_index": 24, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21762,12 +21532,12 @@ "soft_pairs": [ "pull_request:45639|pull_request:45645" ], - "prepared_review_unit_hash": "7fba2bcafca7a77229db8205f474b1f18cd862d457c70267e14010089b8b1ce0", + "prepared_review_unit_hash": "26423451dc28a322605bea862c8af295d042699d3c1bbc44dc3726bb26d812ca", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4590, + "serialized_chars": 4589, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -21775,7 +21545,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4590, + "serialized_chars": 4589, "estimated_input_tokens": 1148, "estimated_eval_tokens": 2552 }, @@ -21784,18 +21554,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both PRs address the same captured_info/xdist collision problem in `src/transformers/testing_utils.py`; PR 45645 is the broader, more complete version because it also updates CI log collection to handle multiple files.", - "confidence": 0.93, + "summary": "Both PRs address xdist-safe handling for captured_info/debug artifact files in the same testing utility path; 45645 appears to subsume the narrower fix in 45639 with additional CI-side log aggregation changes.", + "confidence": 0.89, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45645 is the strongest representative because it contains the core xdist-safe `captured_info` fix plus the related CI/workflow updates needed to preserve and aggregate those logs.", + "canonical_pr_reason": "PR 45645 is the broader, more complete change: it includes the xdist-safe testing_utils fix from 45639 and also updates CI/workflow handling so all captured_info shards are preserved and displayed.", "best_issue_reason": null, - "best_pr_reason": "PR 45645 best captures the cluster\u2019s underlying change: making captured_info artifact handling xdist-safe and ensuring downstream CI/debug log handling still works.", + "best_pr_reason": "PR 45645 is the best representative because it covers the core bug fix plus the surrounding CI/logging adjustments needed for the feature to work end-to-end.", "soft_edge_verdicts": [ { "left": "pull_request:45639", "right": "pull_request:45645", "accept": true, - "reason": "Accepted: both fix the same concrete collision in patched testing debug logs under xdist; 45645 extends the same fix with additional CI/reporting plumbing rather than a different bug." + "reason": "Same underlying xdist collision problem around captured_info output in testing_utils; 45645 largely subsumes 45639\u2019s fix and could plausibly be merged as one PR." } ] }, @@ -21809,7 +21579,7 @@ { "label": "pull_request", "component_index": 25, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21828,12 +21598,12 @@ "pull_request:45670|pull_request:46401", "pull_request:46320|pull_request:46401" ], - "prepared_review_unit_hash": "6278508ae2368fc648feebe13345c2060aca00fd2de9dbedd1a8803d0fa01808", + "prepared_review_unit_hash": "5f9b9753300010c2bf6af320d15660043613069effe48bdf600e609a2e014d56", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7838, + "serialized_chars": 7839, "estimated_input_tokens": 1960, "estimated_eval_tokens": 4176 }, @@ -21841,7 +21611,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 7838, + "serialized_chars": 7839, "estimated_input_tokens": 1960, "estimated_eval_tokens": 4176 }, @@ -21850,30 +21620,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Only the two attn_implementation PRs are superficially related; they target different model families and code paths, so they should remain separate. The AutoModel mapping PR is unrelated.", - "confidence": 0.92, + "summary": "These are three distinct pull requests: one adds an AutoModel mapping for GLM ASR, and the other two add attn_implementation dispatch support for different model families. Despite similar wording, the attn_implementation PRs touch different code paths and model sets, so they should not be merged as duplicates.", + "confidence": 0.93, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46401 is the broadest and most representative of the attn_implementation dispatch work, covering several vision-language model families and their tests/docs. The other PRs are either narrower model-specific variants or unrelated auto-mapping changes.", + "canonical_pr_reason": "PR 46401 is the broadest and most substantial change in the set, covering multiple model families and tests/docs for attn_implementation dispatch support.", "best_issue_reason": null, - "best_pr_reason": "PR 46401 is the best single representative of the cluster theme because it cleanly addresses attn_implementation dispatch across multiple affected model implementations, whereas 46320 is a separate RecurrentGemma-specific change and 45670 is unrelated.", + "best_pr_reason": "PR 46401 is the strongest representative of the cluster because it has the largest scope and the most evidence of a concrete feature implementation, while the other PRs are either narrower or unrelated.", "soft_edge_verdicts": [ { "left": "pull_request:45670", "right": "pull_request:46320", "accept": false, - "reason": "Different changes: 45670 adds GlmAsr entries to auto model mappings, while 46320 implements attn_implementation dispatch for RecurrentGemma. No shared bug or mergeable code-path fix." + "reason": "Different changes: GLM ASR auto-mapping vs RecurrentGemma attn_implementation dispatch support. They do not fix the same concrete code-path problem." }, { "left": "pull_request:45670", "right": "pull_request:46401", "accept": false, - "reason": "Unrelated changes. 45670 is an AutoModel registration tweak for GlmAsr; 46401 adds attn_implementation dispatch support for BLIP-2/InstructBLIP families." + "reason": "Different changes: GLM ASR auto-mapping vs BLIP-2/InstructBLIP attn_implementation dispatch support. No shared bug or mergeable duplicate change." }, { "left": "pull_request:46320", "right": "pull_request:46401", "accept": false, - "reason": "Both are about attn_implementation support, but they fix different model families and different implementation paths. They are too broad and distinct to treat as the same underlying change." + "reason": "Both are attn_implementation-related, but they target different model families and code paths; this is a reusable pattern, not the same underlying bug/change." } ] }, @@ -21887,7 +21657,7 @@ { "label": "pull_request", "component_index": 26, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -21903,40 +21673,40 @@ "soft_pairs": [ "pull_request:45720|pull_request:46259" ], - "prepared_review_unit_hash": "36a53b58b50fe521d92e7f52810f82b3f083b8cebe594ea12481f627cbb47c74", + "prepared_review_unit_hash": "e77c01ab2cba0baff6456163cc9900b254240ab3c5e6150d37957fe33a80dcb7", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7189, - "estimated_input_tokens": 1798, - "estimated_eval_tokens": 3852 + "serialized_chars": 7187, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 7189, - "estimated_input_tokens": 1798, - "estimated_eval_tokens": 3852 + "serialized_chars": 7187, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are two separate linter-rule PRs: #45720 implements part of TRF003 for tuple-return behavior across model forwards, while #46259 adds a different rule (TRF018/TRF016 docs) about dead boolean flags in image/video processors. They overlap only at a broad \"modeling/linter\" level, not at the same code-path or change set.", - "confidence": 0.97, + "summary": "These are two unrelated linter/modeling-rule PRs that happen to overlap on some Transformer model files. PR 45720 implements part of TRF003 for tuple-return behavior, while PR 46259 adds a different TRF018 rule about dead do_* flags in image/video processors; they should not be merged as duplicates.", + "confidence": 0.95, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR as a duplicate target: the PRs address different rules, different APIs, and different affected model families.", + "canonical_pr_reason": "PR 46259 is the cleaner representative because it introduces one clearly scoped rule with a documented spec and targeted code updates, whereas PR 45720 is a broader partial implementation across many model classes.", "best_issue_reason": null, - "best_pr_reason": "#46259 is the better representative of the cluster because it adds a new documented rule with a broad but coherent set of model fixes; however, it is not a duplicate of #45720.", + "best_pr_reason": "PR 46259 is the strongest standalone change in this cluster: it has a single, explicit modeling-rule goal and a cohesive implementation footprint.", "soft_edge_verdicts": [ { "left": "pull_request:45720", "right": "pull_request:46259", "accept": false, - "reason": "Different underlying changes: TRF003 tuple-return forwarding in model forwards vs TRF018/TRF016 dead boolean flags in image/video processor preprocess paths. Shared linter context is too broad to treat as the same PR." + "reason": "Different underlying changes: TRF003 tuple-return handling vs TRF018 dead do_* flag validation. The shared files reflect the same subsystem being linted, not the same bug or change." } ] }, @@ -21950,7 +21720,7 @@ { "label": "pull_request", "component_index": 27, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22016,7 +21786,7 @@ { "label": "pull_request", "component_index": 28, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22032,40 +21802,40 @@ "soft_pairs": [ "pull_request:45932|pull_request:46100" ], - "prepared_review_unit_hash": "a27eecf58198595a0ad78451a8eb3fde349dc144376481ba7f633b4d5bd73f0a", + "prepared_review_unit_hash": "22b9dbf619b5ef92d7d2500ba4769e0aafbabe49cd357078832f9fc8b296e75a", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4301, - "estimated_input_tokens": 1076, - "estimated_eval_tokens": 2408 + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 }, "budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 4301, - "estimated_input_tokens": 1076, - "estimated_eval_tokens": 2408 + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "Both items are unrelated noisy PRs using similar \u201cdo not merge\u201d testing titles. One is a broad CI/model-code change PR, the other is a single-file test mutation PR. They do not appear to fix the same bug or implement the same change.", - "confidence": 0.98, + "summary": "These two PRs are not duplicates: one is a broad CI/modeling change for Grafana/testing infrastructure, while the other is a narrow test-file migration PR with intentional failing assertions. They do not appear to fix the same code-path problem and would not reasonably be merged into one PR.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 45932 is the more substantive representative of the cluster: it touches CI/workflow and multiple model files, whereas 46100 is just a tiny test-only edit. If one PR must stand in for the cluster, 45932 better reflects the broader code change.", + "canonical_pr_reason": "PR 45932 is the more substantial and representative change in the cluster, spanning CI wiring and multiple model files; PR 46100 is a tiny, unrelated test-only migration stub.", "best_issue_reason": null, - "best_pr_reason": "PR 45932 is the best overall PR candidate because it contains concrete implementation changes across code paths and workflow configuration; 46100 is clearly a placeholder/test-mutation PR and is not a good representative of the cluster.", + "best_pr_reason": "PR 45932 is the best representative artifact because it contains the broader actual code/config changes, whereas PR 46100 is just a localized test-file edit.", "soft_edge_verdicts": [ { "left": "pull_request:45932", "right": "pull_request:46100", "accept": false, - "reason": "Reject: these are not the same underlying change. 45932 modifies CI/workflows and several model implementations, while 46100 only edits a ViT test file with deliberate failing assertions. Similar titles and testing intent are not enough to merge them as duplicates." + "reason": "Different underlying changes: 45932 updates CI secrets and multiple model implementations, while 46100 only edits a ViT test file with dummy assertions. They do not share a concrete bug or mergeable fix." } ] }, @@ -22079,7 +21849,7 @@ { "label": "pull_request", "component_index": 29, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22098,12 +21868,12 @@ "pull_request:46073|pull_request:46196", "pull_request:46154|pull_request:46196" ], - "prepared_review_unit_hash": "e70da6f80a8a422a0a9c1bde4635a24077787255ed234c8202d30237e649ea99", + "prepared_review_unit_hash": "ef36305cc5c858681844e79f095bd8f00d75b3899b538200297b529640b06471", "component_budget": { "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 8083, + "serialized_chars": 8082, "estimated_input_tokens": 2021, "estimated_eval_tokens": 4298 }, @@ -22111,7 +21881,7 @@ "node_count": 3, "item_count": 3, "soft_pair_count": 3, - "serialized_chars": 8083, + "serialized_chars": 8082, "estimated_input_tokens": 2021, "estimated_eval_tokens": 4298 }, @@ -22120,30 +21890,30 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are three model-specific PRs adding SDPA dispatch support to different Transformers architectures. They share a pattern and wording, but they target different code paths and should not be treated as duplicates.", - "confidence": 0.95, + "summary": "All three PRs are related by the same feature theme (adding SDPA dispatch support), but they target different model families and distinct code paths, so they should not be deduplicated.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46196 is the broadest and most substantial of the set, covering multiple related speech/model variants with full code and test updates, so it is the best representative if one PR must be chosen for the cluster.", + "canonical_pr_reason": "No PR is a true duplicate of the others. If a representative must be chosen, PR #46196 is the broadest multi-model SDPA dispatch change, but it still covers separate architectures and should remain distinct.", "best_issue_reason": null, - "best_pr_reason": "PR #46196 is the strongest standalone representative of the SDPA-dispatch work because it spans several model implementations and includes corresponding tests, making it the most complete exemplar of the pattern.", + "best_pr_reason": "PR #46196 is the most comprehensive candidate because it spans multiple model implementations and tests, but it is not the same fix as the CTRL or T5 PRs.", "soft_edge_verdicts": [ { "left": "pull_request:46073", "right": "pull_request:46154", "accept": false, - "reason": "Different architectures and code paths (CTRL vs T5); same feature request but not the same concrete fix and not mergeable as one duplicate PR." + "reason": "Both add SDPA dispatch support, but for different model implementations (CTRL vs T5) and different code paths; they are related feature work, not the same concrete change." }, { "left": "pull_request:46073", "right": "pull_request:46196", "accept": false, - "reason": "Both add SDPA support, but they touch unrelated model families (CTRL vs wav2vec2/SeamlessM4T) and different implementations/tests." + "reason": "CTRL SDPA dispatch is a separate model-specific implementation from the wav2vec2/SeamlessM4T changes; too broad to treat as one mergeable fix." }, { "left": "pull_request:46154", "right": "pull_request:46196", "accept": false, - "reason": "T5 and wav2vec2/SeamlessM4T are distinct model code paths; this is a shared pattern, not the same bug/change." + "reason": "T5 SDPA dispatch and the wav2vec2/SeamlessM4T dispatch work affect different architectures and files, so they are not duplicates of the same underlying bug or patch." } ] }, @@ -22157,7 +21927,7 @@ { "label": "pull_request", "component_index": 30, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22173,7 +21943,7 @@ "soft_pairs": [ "pull_request:46126|pull_request:46269" ], - "prepared_review_unit_hash": "00ab7678c100d035d8edb29a583b415fd81f2488669b811dad6656e33904ad4f", + "prepared_review_unit_hash": "6d705601f4b1c10961961768bea7491020d8ff5a072b29601a2ba7d77e749805", "component_budget": { "node_count": 2, "item_count": 2, @@ -22195,18 +21965,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are not duplicates. PR #46269 is a broad distributed-stack base branch with CI and many subsystem changes, while PR #46126 is a narrower EP + Trainer integration patch; they overlap in a few distributed files but do not look like the same concrete fix/change.", - "confidence": 0.88, + "summary": "Both items are pull requests in the distributed stack, but they are not the same concrete change. PR #46126 is a focused EP + Trainer integration feature, while PR #46269 is a broad distributed branch base touching many unrelated areas. The filename overlap is not enough to treat them as duplicates.", + "confidence": 0.92, "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46269 is the broader umbrella change set and best represents the cluster\u2019s shared distributed branch work; PR #46126 is a narrower feature integration layered on top of that space.", + "canonical_pr_reason": "No true duplicate PR pair here; if one needs to represent the cluster, #46126 is the more focused and self-contained change, whereas #46269 is a broad branch-base PR rather than the same fix.", "best_issue_reason": null, - "best_pr_reason": "PR #46269 is the best single representative because it spans the widest set of related distributed changes and appears to be the base branch that other work builds on.", + "best_pr_reason": "#46126 is the better representative because it has a clear, specific feature goal and a tighter code-path scope. #46269 is too broad and mixed to serve as a canonical duplicate target.", "soft_edge_verdicts": [ { "left": "pull_request:46126", "right": "pull_request:46269", "accept": false, - "reason": "Overlap in distributed files, but the PRs target different scopes: #46126 is EP/Trainer integration, while #46269 is a broad distributed branch base with CI and many unrelated component updates. They are not the same concrete mergeable fix." + "reason": "Reject: overlapping distributed filenames, but different scope and intent. #46126 adds EP/Trainer integration; #46269 is a broad distributed branch base. They do not appear to fix the same concrete bug or implement the same change." } ] }, @@ -22220,7 +21990,7 @@ { "label": "pull_request", "component_index": 31, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22228,48 +21998,150 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-46138-2", + "cluster_id": "cluster-46138-6", "nodes": [ "pull_request:46138", - "pull_request:46432" + "pull_request:46432", + "pull_request:46501", + "pull_request:46502", + "pull_request:46503", + "pull_request:46504" ], "soft_pairs": [ - "pull_request:46138|pull_request:46432" + "pull_request:46138|pull_request:46504", + "pull_request:46138|pull_request:46432", + "pull_request:46138|pull_request:46503", + "pull_request:46138|pull_request:46501", + "pull_request:46138|pull_request:46502", + "pull_request:46432|pull_request:46503", + "pull_request:46432|pull_request:46501", + "pull_request:46432|pull_request:46502", + "pull_request:46432|pull_request:46504", + "pull_request:46501|pull_request:46503", + "pull_request:46501|pull_request:46502", + "pull_request:46501|pull_request:46504", + "pull_request:46502|pull_request:46503", + "pull_request:46502|pull_request:46504", + "pull_request:46503|pull_request:46504" ], - "prepared_review_unit_hash": "ed9a73ee70a03e4235dc562a8412b6bb3a1f6c2393c3c3d13c6676eaa39de5b0", + "prepared_review_unit_hash": "e478cef77dea0d7346b24f131693ee5af22e5a3db8b3f454acd2b0c2a0dc18ee", "component_budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 3863, - "estimated_input_tokens": 966, - "estimated_eval_tokens": 2188 + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14229, + "estimated_input_tokens": 3558, + "estimated_eval_tokens": 7372 }, "budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 3863, - "estimated_input_tokens": 966, - "estimated_eval_tokens": 2188 + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14229, + "estimated_input_tokens": 3558, + "estimated_eval_tokens": 7372 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are unrelated GitHub Actions workflow changes: one updates self-comment CI to pass comment text via stdin, the other hardens ssh-runner by pinning a downloaded helper script and clearing HF_TOKEN before SSH.", - "confidence": 0.96, + "summary": "All six pull requests are small, unrelated workflow/CI maintenance changes. They share only superficial similarity in title and file type, not the same underlying bug or change, so none should be clustered as duplicates.", + "confidence": 0.97, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46432 is the stronger canonical representative because it makes a more substantive workflow/security change (pinning a remote script and unsetting a sensitive token), while 46138 is a smaller command-invocation cleanup.", + "canonical_pr_reason": null, "best_issue_reason": null, - "best_pr_reason": "PR 46432 is the best PR candidate overall for the cluster because it is the more impactful and self-contained change in this set.", + "best_pr_reason": null, "soft_edge_verdicts": [ { "left": "pull_request:46138", "right": "pull_request:46432", "accept": false, - "reason": "Different workflows and different underlying problems. 46138 only changes how a comment is piped into a helper script in self-comment CI; 46432 hardens ssh-runner by pinning a downloaded script and clearing HF_TOKEN. They are not the same code-path fix and would not plausibly merge into one PR." + "reason": "Different workflows and different intents: one changes self-comment CI stdin handling, the other hardens SSH runner download/secret handling." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46501", + "accept": false, + "reason": "Unrelated workflow targets; anti-slop permissions change is not the same code-path as self-comment CI command piping." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46502", + "accept": false, + "reason": "Different automation job and behavior; assign-reviewers checkout ref fix is unrelated to self-comment CI." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46503", + "accept": false, + "reason": "Benchmark workflow trigger expansion is a separate change from self-comment CI input handling." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Build-image trigger parsing fix is a different workflow and problem than self-comment CI stdin passing." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "Both touch CI workflows, but one is SSH/security hardening and the other is benchmark trigger logic; no shared bug or mergeable combined fix." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "Permissions change in anti-slop is unrelated to SSH runner secret cleanup and checksum pinning." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Assign-reviewers checkout ref fix is a distinct workflow issue, not the SSH runner hardening change." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46504", + "accept": false, + "reason": "Build-ci-docker-images command quoting fix is unrelated to SSH runner download/secret handling." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Different workflows and different behavior changes; permissions scoping vs benchmark dispatch support." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Anti-slop permission placement and assign-reviewers checkout ref are unrelated fixes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "No common underlying bug: anti-slop permissions vs build-ci-docker-images shell quoting." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Assign-reviewers checkout ref fix and benchmark workflow dispatch enablement address separate jobs and problems." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Distinct workflows with different failure modes; no same concrete code-path problem." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark dispatch support and build-ci-docker-images quoting fix are independent workflow changes, not one duplicate fix." } ] }, @@ -22283,7 +22155,7 @@ { "label": "pull_request", "component_index": 32, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22299,7 +22171,7 @@ "soft_pairs": [ "pull_request:46152|pull_request:46207" ], - "prepared_review_unit_hash": "5313e57188fd32fd048bc38f8dd8d9b5af54bb0de9c4648a3e88792fff4fb557", + "prepared_review_unit_hash": "d7d733bf9bc2c6037704fa3e05b8b0c1367ab93f4dca3fa22b013af54b18f547", "component_budget": { "node_count": 2, "item_count": 2, @@ -22321,18 +22193,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These are related source-reading hardening PRs, but they fix different code paths in different modules, so they are not duplicates.", - "confidence": 0.89, + "summary": "Two open PRs both replace file-opening source introspection, but they target different modules and different behaviors: one fixes package-resource resolution in import_utils, the other changes attention-implementation detection in modeling_utils. They are related in style, not duplicates.", + "confidence": 0.94, "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46207 is the better canonical PR representative: it targets a specific `_can_set_attn_implementation` path in `modeling_utils.py` and includes tests, making the change clearer and better-scoped.", + "canonical_pr_reason": "No clear canonical PR: the two PRs address separate code paths and bugs rather than one shared change.", "best_issue_reason": null, - "best_pr_reason": "PR 46207 is the strongest overall representative of the cluster because it is a concrete, test-backed fix to one code path, whereas 46152 is a separate utility-level change in another module.", + "best_pr_reason": "PR 46207 is the more concrete and self-contained fix, with a specific behavior change in modeling_utils and accompanying tests.", "soft_edge_verdicts": [ { "left": "pull_request:46152", "right": "pull_request:46207", "accept": false, - "reason": "Both replace file-open source inspection with a safer source lookup, but they affect different modules and different behaviors (`import_utils.py` vs `modeling_utils.py`). They are thematically similar, not the same underlying bug, and would not naturally merge into one PR." + "reason": "Both avoid open()-based source access, but they fix different components and different failure modes; they would not plausibly be merged into one PR." } ] }, @@ -22346,7 +22218,7 @@ { "label": "pull_request", "component_index": 33, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22354,93 +22226,62 @@ "source": "llm", "cache_hit": false, "model": "gpt-5.4-mini?service_tier=flex", - "cluster_id": "cluster-46173-6", + "cluster_id": "cluster-46173-3", "nodes": [ "pull_request:46173", - "pull_request:46343", "pull_request:46436", - "pull_request:46474", - "pull_request:46475", - "pull_request:46481" + "pull_request:46474" ], "soft_pairs": [ - "pull_request:46436|pull_request:46481", "pull_request:46436|pull_request:46474", - "pull_request:46474|pull_request:46481", - "pull_request:46474|pull_request:46475", - "pull_request:46173|pull_request:46474", - "pull_request:46343|pull_request:46475" + "pull_request:46173|pull_request:46474" ], - "prepared_review_unit_hash": "93038bf9543d9f36cccc6a4d95e0e9f122f15df3a65a56a2e3eaf8ae8da21aa9", + "prepared_review_unit_hash": "9060d9b3132af9e84e0b2842c60ceb5c437a219f67fd887ac2ff160e19813c64", "component_budget": { - "node_count": 6, - "item_count": 6, - "soft_pair_count": 6, - "serialized_chars": 14073, - "estimated_input_tokens": 3519, - "estimated_eval_tokens": 7294 + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 }, "budget": { - "node_count": 6, - "item_count": 6, - "soft_pair_count": 6, - "serialized_chars": 14073, - "estimated_input_tokens": 3519, - "estimated_eval_tokens": 7294 + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 }, "overflow_policy": "truncate_then_skip", "trimmed": false, "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "This cluster is not one duplicate set; it splits into three separate PR pairs: a StaticLayer type-hint fix, a Falcon-Mamba kernel-warning wording fix, and a static-cache max-length config fix. I only accept soft edges within each pair.", - "confidence": 0.95, + "summary": "Two PRs are effectively duplicates of the same StaticLayer type-hint fix for #45987, while the static-cache max-len PR is unrelated and should stay separate.", + "confidence": 0.98, "canonical_issue_reason": null, - "canonical_pr_reason": "No single canonical PR spans the whole cluster because the items are three unrelated fixes; the nearest duplicate pairwise representatives are 46474, 46475, and 46481 for their respective bugs.", + "canonical_pr_reason": "PR 46173 is the earliest instance of the exact same cache_utils type-annotation change for #45987, so it is the best anchor for the duplicate cluster.", "best_issue_reason": null, - "best_pr_reason": "46474 is the strongest single representative among the duplicate pairs because it matches the cache_utils type-hint fix exactly and includes tests, but it is only representative of one subcluster.", + "best_pr_reason": "PR 46173 is the strongest representative because it is the original, exact fix for the shared issue target and matches the later duplicate nearly verbatim.", "soft_edge_verdicts": [ - { - "left": "pull_request:46173", - "right": "pull_request:46474", - "accept": true, - "reason": "Same underlying StaticLayer get_seq_length type-hint fix in cache_utils; 46474 is just the test-backed version of the same change." - }, - { - "left": "pull_request:46343", - "right": "pull_request:46475", - "accept": true, - "reason": "Same Falcon-Mamba fast-path warning clarification on the same code path; both are the same wording fix, with 46475 adding tests." - }, { "left": "pull_request:46436", "right": "pull_request:46474", "accept": false, - "reason": "Different bugs and code paths: generation cache max-length handling vs StaticLayer return-type annotation." - }, - { - "left": "pull_request:46474", - "right": "pull_request:46475", - "accept": false, - "reason": "Unrelated subsystems and fixes: cache_utils type hint versus Falcon-Mamba warning text." - }, - { - "left": "pull_request:46474", - "right": "pull_request:46481", - "accept": false, - "reason": "Different changes in generation/cache handling; shared test file is not enough to make them the same bug." + "reason": "These fix different problems: 46436 changes generation cache sizing from config, while 46474 only widens StaticLayer.get_seq_length type hints. Same broad subsystem, but not the same bug or mergeable change set." }, { - "left": "pull_request:46436", - "right": "pull_request:46481", + "left": "pull_request:46173", + "right": "pull_request:46474", "accept": true, - "reason": "Same static-cache max-length config bug in generation/utils.py; both PRs implement the same fix and could be merged into one." + "reason": "They are the same underlying fix: both edit src/transformers/cache_utils.py with the same get_seq_length return-type annotation change and the same issue target #45987." } ] }, "evaluator_result": { "accept": true, - "feedback": "The summary is grounded in the packet: the six PRs cleanly split into three issue-target subclusters (45987, 46234, 46424), and the accepted soft edges are limited to PRs that are plausibly the same bugfix and mergeable. The rejected edges are conservative and correctly avoid cross-bug linkage." + "feedback": "The summary is grounded in the packet: PRs 46173 and 46474 have the same issue target (#45987) and the same cache_utils type-annotation diff, so treating them as duplicates is conservative and supported. The 46436 vs 46474 verdict is also appropriately conservative because the diffs address different features/problems and are not mergeable as one bugfix PR." }, "evaluator_used": true, "retried": false, @@ -22451,7 +22292,7 @@ { "label": "pull_request", "component_index": 34, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22514,85 +22355,7 @@ { "label": "pull_request", "component_index": 35, - "component_count": 38, - "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-46337-3", - "nodes": [ - "pull_request:46337", - "pull_request:46362", - "pull_request:46368" - ], - "soft_pairs": [ - "pull_request:46337|pull_request:46368", - "pull_request:46337|pull_request:46362", - "pull_request:46362|pull_request:46368" - ], - "prepared_review_unit_hash": "bbe8f414c8d8bb7b9dd0e8d1536cfc6fa3ebe80607dbc19f3e652d1aeebf5740", - "component_budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 5897, - "estimated_input_tokens": 1475, - "estimated_eval_tokens": 3206 - }, - "budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 5897, - "estimated_input_tokens": 1475, - "estimated_eval_tokens": 3206 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": false, - "analyst_result": { - "summary": "These are three separate test-fix PRs, not duplicates. The two quant_cache skips are model-specific workarounds for different cache implementations, and the Cohere2Moe PR is a different device-map/test-setup fix entirely.", - "confidence": 0.91, - "canonical_issue_reason": null, - "canonical_pr_reason": null, - "best_issue_reason": null, - "best_pr_reason": "No single PR cleanly subsumes the others. If forced to pick a representative, 46337 is the broadest fix, but it targets a different failure mode than the quant_cache skips.", - "soft_edge_verdicts": [ - { - "left": "pull_request:46337", - "right": "pull_request:46368", - "accept": false, - "reason": "Different bugs: Cohere2Moe device-map/setup failure vs NemotronH quantized-cache incompatibility. Same broad test area, but not the same concrete change." - }, - { - "left": "pull_request:46337", - "right": "pull_request:46362", - "accept": false, - "reason": "Different failure modes and files; one fixes Cohere2Moe device placement, the other skips a DeepSeek V4 quant-cache test." - }, - { - "left": "pull_request:46362", - "right": "pull_request:46368", - "accept": false, - "reason": "Both skip the same test name pattern, but for different model-specific cache implementations and distinct incompatibilities, so they are not one mergeable 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": 36, - "component_count": 38, + "component_count": 35, "review_unit_index": 1, "review_unit_count": 1, "status": "reviewed", @@ -22608,12 +22371,12 @@ "soft_pairs": [ "pull_request:46376|pull_request:46417" ], - "prepared_review_unit_hash": "715937e8046c740550f820152b1d5f5491a8b170168d4234d72fde739686f492", + "prepared_review_unit_hash": "645314fa075543245856d8a30a2f2e83744a37df47f010f857e4bc7e8909e10b", "component_budget": { "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5959, + "serialized_chars": 5960, "estimated_input_tokens": 1490, "estimated_eval_tokens": 3236 }, @@ -22621,7 +22384,7 @@ "node_count": 2, "item_count": 2, "soft_pair_count": 1, - "serialized_chars": 5959, + "serialized_chars": 5960, "estimated_input_tokens": 1490, "estimated_eval_tokens": 3236 }, @@ -22630,159 +22393,18 @@ "aggressively_trimmed": false, "split": false, "analyst_result": { - "summary": "These pull requests are not duplicates: #46376 is a broad draft for multimodal continuous batching infrastructure, while #46417 adds the HunYuan VL model, docs, and tests. They touch different code paths and should not be merged as one change.", - "confidence": 0.95, + "summary": "The two pull requests are unrelated: one adds multimodal support to continuous batching, while the other adds a new HunYuan VL model implementation and docs. They do not look like the same underlying fix or change set.", + "confidence": 0.96, "canonical_issue_reason": null, - "canonical_pr_reason": "#46417 is the cleaner canonical representative: it is a self-contained model addition with focused scope, docs, and tests, whereas #46376 is a broad draft infrastructure refactor.", + "canonical_pr_reason": "PR 46417 is the better representative artifact because it is a concrete, self-contained model addition with code, docs, and tests; PR 46376 is a draft, broad infrastructure feature and not a duplicate of it.", "best_issue_reason": null, - "best_pr_reason": "#46417 is the best PR to represent this cluster because it is concrete and cohesive; #46376 is unrelated in function and still a draft multimodal batching overhaul.", + "best_pr_reason": "PR 46417 is the strongest standalone PR in the cluster: it has a clearly defined scope and substantial implementation coverage. PR 46376 is broader, still draft, and targets a different subsystem.", "soft_edge_verdicts": [ { "left": "pull_request:46376", "right": "pull_request:46417", "accept": false, - "reason": "Different underlying changes: continuous-batching multimodal support vs. a new HunYuan VL model implementation. No shared concrete bug or mergeable 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": 37, - "component_count": 38, - "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-46465-2", - "nodes": [ - "pull_request:46465", - "pull_request:46470" - ], - "soft_pairs": [ - "pull_request:46465|pull_request:46470" - ], - "prepared_review_unit_hash": "2d028eb1ee5c9984d03de8b7fe0120f92f53c9aa09665e236207f7f1d4160ba8", - "component_budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 4986, - "estimated_input_tokens": 1247, - "estimated_eval_tokens": 2750 - }, - "budget": { - "node_count": 2, - "item_count": 2, - "soft_pair_count": 1, - "serialized_chars": 4986, - "estimated_input_tokens": 1247, - "estimated_eval_tokens": 2750 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": false, - "analyst_result": { - "summary": "Both PRs target the same decoder input-validation error-message area in T5-family models, but they implement different text fixes. One broadens/rewords the message across more model files, while the other narrowly corrects the decoder prefix formatting and adds tests.", - "confidence": 0.74, - "canonical_issue_reason": null, - "canonical_pr_reason": "PR #46470 is the cleaner anchor: it directly fixes the decoder validation message formatting in the affected T5/MT5/uMT5 code paths and includes tests.", - "best_issue_reason": null, - "best_pr_reason": "PR #46470 is the best overall candidate because it is focused, tested, and matches the concrete error-message bug without extra wording changes.", - "soft_edge_verdicts": [ - { - "left": "pull_request:46465", - "right": "pull_request:46470", - "accept": false, - "reason": "Shared issue target aside, these are not the same concrete change: #46465 adds a new explanatory sentence and broader wording changes across more model files, while #46470 only changes the decoder prefix formatting and test coverage." - } - ] - }, - "evaluator_result": null, - "evaluator_used": false, - "retried": false, - "accepted_nontrivial_soft_edge": false, - "error_kind": null, - "error_message": null - }, - { - "label": "pull_request", - "component_index": 38, - "component_count": 38, - "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-46483-3", - "nodes": [ - "pull_request:46483", - "pull_request:46484", - "pull_request:46485" - ], - "soft_pairs": [ - "pull_request:46483|pull_request:46484", - "pull_request:46483|pull_request:46485", - "pull_request:46484|pull_request:46485" - ], - "prepared_review_unit_hash": "f208880f5b83aaa4cbce62da46a151d5ab1261300cfdd751396b3ac276b58d76", - "component_budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 8159, - "estimated_input_tokens": 2040, - "estimated_eval_tokens": 4336 - }, - "budget": { - "node_count": 3, - "item_count": 3, - "soft_pair_count": 3, - "serialized_chars": 8159, - "estimated_input_tokens": 2040, - "estimated_eval_tokens": 4336 - }, - "overflow_policy": "truncate_then_skip", - "trimmed": false, - "aggressively_trimmed": false, - "split": false, - "analyst_result": { - "summary": "These PRs are not duplicates: one hardens `trust_remote_code` handling for local custom generation, while the other two add separate allocation guards in different model code paths. The two memory-safety PRs are similar in theme but fix different concrete bugs.", - "confidence": 0.92, - "canonical_issue_reason": null, - "canonical_pr_reason": "PR 46484 is the best representative of the memory-safety fixes: it applies a concrete allocation bound across many positional-embedding implementations and includes test coverage.", - "best_issue_reason": null, - "best_pr_reason": "PR 46484 is the strongest standalone change in the cluster because it is broad, specific, and directly addresses an eager OOM risk with a clear guard and test.", - "soft_edge_verdicts": [ - { - "left": "pull_request:46483", - "right": "pull_request:46484", - "accept": false, - "reason": "Different bug classes and code paths: remote-code trust gating vs. sinusoidal embedding allocation limits. Similar safety theme, but not the same fix." - }, - { - "left": "pull_request:46483", - "right": "pull_request:46485", - "accept": false, - "reason": "Unrelated fixes: custom generation loading security vs. causal-mask size guarding. They do not look mergeable into one PR." - }, - { - "left": "pull_request:46484", - "right": "pull_request:46485", - "accept": false, - "reason": "Both prevent OOMs, but in different mechanisms and model paths (positional embeddings vs. causal masks). They share a theme, not the same underlying bug or change." + "reason": "Different work items and code paths: continuous batching multimodal infrastructure versus adding the HunYuan VL model package, configs, processors, and tests. They are not plausibly mergeable into one PR as duplicates." } ] }, diff --git a/analysis/current/manifest.json b/analysis/current/manifest.json index 72faa8e73062dc4c8cc67af8adcbab1318a12462..200c78e9dbacfbc1ef34be42910907dd241f07bd 100644 --- a/analysis/current/manifest.json +++ b/analysis/current/manifest.json @@ -1,8 +1,8 @@ { - "analysis_id": "hybrid-model-20260608t120020z", + "analysis_id": "hybrid-model-20260608t180018z", "archived_artifacts": { - "hybrid": "snapshots/20260608T120020Z/analysis-runs/hybrid-model-20260608t120020z/analysis-report-hybrid.json", - "hybrid_reviews": "snapshots/20260608T120020Z/analysis-runs/hybrid-model-20260608t120020z/analysis-report-hybrid.llm-reviews.json" + "hybrid": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json" }, "artifacts": { "hybrid": "analysis/current/analysis-report-hybrid.json", @@ -10,9 +10,9 @@ }, "channel": "canonical", "model": null, - "published_at": "2026-06-08T12:17:15Z", + "published_at": "2026-06-08T18:11:35Z", "repo": "huggingface/transformers", "schema_version": 1, - "snapshot_id": "20260608T120020Z", + "snapshot_id": "20260608T180018Z", "variant": "hybrid" } diff --git a/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json new file mode 100644 index 0000000000000000000000000000000000000000..f01dd714caf519ba0073d6aca35173a05cce2f90 --- /dev/null +++ b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json @@ -0,0 +1,2506 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260608T180018Z", + "generated_at": "2026-06-08T18:11:33Z", + "evidence_quality": "full", + "llm_enrichment": true, + "meta_bugs": [ + { + "cluster_id": "cluster-43979-11", + "summary": "Cluster of 1 issues and 10 PRs centered on issue #43979.", + "status": "open", + "confidence": 0.8, + "canonical_issue_number": 43979, + "canonical_pr_number": 44007, + "issue_numbers": [ + 43979 + ], + "pr_numbers": [ + 43996, + 44007, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target", + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 43996, + "right_pr_number": 44007, + "code_similarity": 0.179, + "size_similarity": 0.576, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.429, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44013, + "code_similarity": 0.122, + "size_similarity": 0.318, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.392, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44044, + "code_similarity": 0.245, + "size_similarity": 0.864, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.479, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44066, + "code_similarity": 0.225, + "size_similarity": 0.818, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.408, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44072, + "code_similarity": 0.14, + "size_similarity": 0.303, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.528, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44085, + "code_similarity": 0.216, + "size_similarity": 0.783, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.398, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44129, + "code_similarity": 0.163, + "size_similarity": 0.643, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.229, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44154, + "code_similarity": 0.153, + "size_similarity": 0.535, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.31, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 43996, + "right_pr_number": 44722, + "code_similarity": 0.225, + "size_similarity": 0.848, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.368, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44013, + "code_similarity": 0.19, + "size_similarity": 0.553, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.531, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44044, + "code_similarity": 0.186, + "size_similarity": 0.667, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.354, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44066, + "code_similarity": 0.188, + "size_similarity": 0.704, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.315, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44072, + "code_similarity": 0.212, + "size_similarity": 0.526, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.708, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44085, + "code_similarity": 0.195, + "size_similarity": 0.735, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.318, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44129, + "code_similarity": 0.103, + "size_similarity": 0.37, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.191, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44154, + "code_similarity": 0.238, + "size_similarity": 0.93, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.344, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44007, + "right_pr_number": 44722, + "code_similarity": 0.178, + "size_similarity": 0.679, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.28, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44044, + "code_similarity": 0.126, + "size_similarity": 0.368, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.351, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44066, + "code_similarity": 0.127, + "size_similarity": 0.389, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.325, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44072, + "code_similarity": 0.29, + "size_similarity": 0.952, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.667, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44085, + "code_similarity": 0.131, + "size_similarity": 0.406, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.329, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44129, + "code_similarity": 0.07, + "size_similarity": 0.205, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.192, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44154, + "code_similarity": 0.177, + "size_similarity": 0.594, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.389, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44013, + "right_pr_number": 44722, + "code_similarity": 0.118, + "size_similarity": 0.375, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.287, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44066, + "code_similarity": 0.25, + "size_similarity": 0.947, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.404, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44072, + "code_similarity": 0.136, + "size_similarity": 0.351, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.442, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44085, + "code_similarity": 0.24, + "size_similarity": 0.906, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.394, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44129, + "code_similarity": 0.147, + "size_similarity": 0.555, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.243, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44154, + "code_similarity": 0.17, + "size_similarity": 0.62, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.306, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44044, + "right_pr_number": 44722, + "code_similarity": 0.257, + "size_similarity": 0.982, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.402, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44072, + "code_similarity": 0.133, + "size_similarity": 0.37, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.393, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44085, + "code_similarity": 0.763, + "size_similarity": 0.957, + "file_overlap": 0.5, + "area_overlap": 0.825, + "patch_similarity": 0.887, + "shared_filenames": [ + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 41 + ], + [ + 174, + 181 + ], + [ + 250, + 257 + ], + [ + 398, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 466 + ], + [ + 471, + 488 + ], + [ + 518, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 606 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ], + "right_ranges": [ + [ + 33, + 40 + ], + [ + 173, + 180 + ], + [ + 249, + 256 + ], + [ + 397, + 405 + ], + [ + 408, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 465 + ], + [ + 470, + 489 + ], + [ + 517, + 539 + ], + [ + 553, + 559 + ], + [ + 565, + 574 + ], + [ + 579, + 597 + ], + [ + 600, + 611 + ], + [ + 633, + 639 + ], + [ + 645, + 652 + ], + [ + 657, + 673 + ], + [ + 716, + 728 + ], + [ + 737, + 743 + ], + [ + 748, + 771 + ], + [ + 789, + 794 + ] + ] + } + ] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44129, + "code_similarity": 0.145, + "size_similarity": 0.526, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.263, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44154, + "code_similarity": 0.174, + "size_similarity": 0.654, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.286, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44066, + "right_pr_number": 44722, + "code_similarity": 0.907, + "size_similarity": 0.964, + "file_overlap": 1.0, + "area_overlap": 0.808, + "patch_similarity": 0.874, + "shared_filenames": [ + "src/transformers/models/codegen/modeling_codegen.py", + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/codegen/modeling_codegen.py", + "left_ranges": [ + [ + 245, + 266 + ] + ], + "right_ranges": [ + [ + 228, + 234 + ] + ] + }, + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 41 + ], + [ + 174, + 181 + ], + [ + 250, + 257 + ], + [ + 398, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 466 + ], + [ + 471, + 488 + ], + [ + 518, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 606 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ], + "right_ranges": [ + [ + 33, + 42 + ], + [ + 175, + 182 + ], + [ + 251, + 258 + ], + [ + 399, + 420 + ], + [ + 426, + 435 + ], + [ + 460, + 467 + ], + [ + 472, + 489 + ], + [ + 519, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 612 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 663 + ], + [ + 666, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ] + } + ] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44085, + "code_similarity": 0.137, + "size_similarity": 0.387, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.398, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44129, + "code_similarity": 0.074, + "size_similarity": 0.195, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.231, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44154, + "code_similarity": 0.175, + "size_similarity": 0.566, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.414, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44072, + "right_pr_number": 44722, + "code_similarity": 0.124, + "size_similarity": 0.357, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.347, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44129, + "code_similarity": 0.141, + "size_similarity": 0.503, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.272, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44154, + "code_similarity": 0.18, + "size_similarity": 0.684, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.289, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44085, + "right_pr_number": 44722, + "code_similarity": 0.728, + "size_similarity": 0.923, + "file_overlap": 0.5, + "area_overlap": 0.791, + "patch_similarity": 0.78, + "shared_filenames": [ + "src/transformers/models/gptj/modeling_gptj.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gptj/modeling_gptj.py", + "left_ranges": [ + [ + 33, + 40 + ], + [ + 173, + 180 + ], + [ + 249, + 256 + ], + [ + 397, + 405 + ], + [ + 408, + 419 + ], + [ + 425, + 434 + ], + [ + 459, + 465 + ], + [ + 470, + 489 + ], + [ + 517, + 539 + ], + [ + 553, + 559 + ], + [ + 565, + 574 + ], + [ + 579, + 597 + ], + [ + 600, + 611 + ], + [ + 633, + 639 + ], + [ + 645, + 652 + ], + [ + 657, + 673 + ], + [ + 716, + 728 + ], + [ + 737, + 743 + ], + [ + 748, + 771 + ], + [ + 789, + 794 + ] + ], + "right_ranges": [ + [ + 33, + 42 + ], + [ + 175, + 182 + ], + [ + 251, + 258 + ], + [ + 399, + 420 + ], + [ + 426, + 435 + ], + [ + 460, + 467 + ], + [ + 472, + 489 + ], + [ + 519, + 540 + ], + [ + 554, + 560 + ], + [ + 566, + 575 + ], + [ + 580, + 598 + ], + [ + 601, + 612 + ], + [ + 634, + 640 + ], + [ + 646, + 653 + ], + [ + 658, + 663 + ], + [ + 666, + 674 + ], + [ + 717, + 722 + ], + [ + 738, + 744 + ], + [ + 749, + 772 + ], + [ + 790, + 795 + ] + ] + } + ] + }, + { + "left_pr_number": 44129, + "right_pr_number": 44154, + "code_similarity": 0.099, + "size_similarity": 0.344, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.199, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44129, + "right_pr_number": 44722, + "code_similarity": 0.146, + "size_similarity": 0.545, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.247, + "shared_filenames": [], + "shared_file_areas": [] + }, + { + "left_pr_number": 44154, + "right_pr_number": 44722, + "code_similarity": 0.164, + "size_similarity": 0.631, + "file_overlap": 0.0, + "area_overlap": 0.0, + "patch_similarity": 0.255, + "shared_filenames": [], + "shared_file_areas": [] + } + ] + }, + { + "cluster_id": "cluster-41211-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41211.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41211, + "canonical_pr_number": 44339, + "issue_numbers": [ + 41211 + ], + "pr_numbers": [ + 41356, + 44339 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41356, + "right_pr_number": 44339, + "code_similarity": 0.155, + "size_similarity": 0.094, + "file_overlap": 0.273, + "area_overlap": 0.078, + "patch_similarity": 0.183, + "shared_filenames": [ + "docs/source/en/model_doc/deimv2.md", + "src/transformers/models/deimv2/__init__.py", + "src/transformers/models/deimv2/configuration_deimv2.py", + "src/transformers/models/deimv2/modeling_deimv2.py", + "tests/models/deimv2/__init__.py", + "tests/models/deimv2/test_modeling_deimv2.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/model_doc/deimv2.md", + "left_ranges": [ + [ + 1, + 132 + ] + ], + "right_ranges": [ + [ + 1, + 65 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/__init__.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/deimv2/configuration_deimv2.py", + "left_ranges": [ + [ + 1, + 74 + ] + ], + "right_ranges": [ + [ + 1, + 266 + ] + ] + }, + { + "filename": "tests/models/deimv2/test_modeling_deimv2.py", + "left_ranges": [ + [ + 1, + 15 + ] + ], + "right_ranges": [ + [ + 1, + 1734 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45363-3", + "summary": "Cluster of 3 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 46339, + "issue_numbers": [], + "pr_numbers": [ + 45363, + 45666, + 46339 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 45363, + "right_pr_number": 45666, + "code_similarity": 0.392, + "size_similarity": 0.298, + "file_overlap": 0.571, + "area_overlap": 0.329, + "patch_similarity": 0.306, + "shared_filenames": [ + "fused_qwen_example.py", + "src/transformers/integrations/__init__.py", + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py" + ], + "shared_file_areas": [ + { + "filename": "fused_qwen_example.py", + "left_ranges": [ + [ + 1, + 30 + ] + ], + "right_ranges": [ + [ + 1, + 49 + ] + ] + }, + { + "filename": "src/transformers/integrations/__init__.py", + "left_ranges": [ + [ + 67, + 81 + ], + [ + 226, + 240 + ] + ], + "right_ranges": [ + [ + 67, + 73 + ], + [ + 222, + 228 + ] + ] + }, + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 17 + ], + [ + 22, + 35 + ], + [ + 505, + 716 + ] + ], + "right_ranges": [ + [ + 11, + 53 + ], + [ + 286, + 295 + ], + [ + 521, + 1077 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 3738, + 3747 + ] + ], + "right_ranges": [ + [ + 1330, + 1338 + ], + [ + 3763, + 3789 + ], + [ + 4223, + 4234 + ], + [ + 4628, + 4641 + ] + ] + } + ] + }, + { + "left_pr_number": 45363, + "right_pr_number": 46339, + "code_similarity": 0.506, + "size_similarity": 0.925, + "file_overlap": 0.286, + "area_overlap": 0.545, + "patch_similarity": 0.297, + "shared_filenames": [ + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 17 + ], + [ + 22, + 35 + ], + [ + 505, + 716 + ] + ], + "right_ranges": [ + [ + 14, + 52 + ], + [ + 285, + 294 + ], + [ + 521, + 685 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 3738, + 3747 + ] + ], + "right_ranges": [ + [ + 1346, + 1354 + ], + [ + 3785, + 3811 + ], + [ + 4245, + 4256 + ], + [ + 4650, + 4663 + ] + ] + } + ] + }, + { + "left_pr_number": 45666, + "right_pr_number": 46339, + "code_similarity": 0.399, + "size_similarity": 0.322, + "file_overlap": 0.667, + "area_overlap": 0.247, + "patch_similarity": 0.318, + "shared_filenames": [ + "src/transformers/core_model_loading.py", + "src/transformers/integrations/hub_kernels.py", + "src/transformers/modeling_utils.py", + "tests/kernels/test_kernels.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/core_model_loading.py", + "left_ranges": [ + [ + 19, + 25 + ], + [ + 80, + 86 + ] + ], + "right_ranges": [ + [ + 19, + 25 + ], + [ + 80, + 86 + ] + ] + }, + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 11, + 53 + ], + [ + 286, + 295 + ], + [ + 521, + 1077 + ] + ], + "right_ranges": [ + [ + 14, + 52 + ], + [ + 285, + 294 + ], + [ + 521, + 685 + ] + ] + }, + { + "filename": "src/transformers/modeling_utils.py", + "left_ranges": [ + [ + 1330, + 1338 + ], + [ + 3763, + 3789 + ], + [ + 4223, + 4234 + ], + [ + 4628, + 4641 + ] + ], + "right_ranges": [ + [ + 1346, + 1354 + ], + [ + 3785, + 3811 + ], + [ + 4245, + 4256 + ], + [ + 4650, + 4663 + ] + ] + }, + { + "filename": "tests/kernels/test_kernels.py", + "left_ranges": [ + [ + 19, + 54 + ], + [ + 228, + 281 + ], + [ + 294, + 317 + ], + [ + 605, + 995 + ] + ], + "right_ranges": [ + [ + 21, + 28 + ], + [ + 33, + 39 + ], + [ + 90, + 103 + ], + [ + 217, + 323 + ], + [ + 336, + 359 + ] + ] + } + ] + } + ] + }, + { + "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-41115-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #41115.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 41115, + "canonical_pr_number": 45613, + "issue_numbers": [ + 41115 + ], + "pr_numbers": [ + 41116, + 45613 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 41116, + "right_pr_number": 45613, + "code_similarity": 0.797, + "size_similarity": 0.885, + "file_overlap": 0.75, + "area_overlap": 0.831, + "patch_similarity": 0.693, + "shared_filenames": [ + "docs/source/en/_toctree.yml", + "docs/source/en/model_doc/minicpm3.md", + "src/transformers/models/auto/modeling_auto.py", + "src/transformers/models/minicpm3/__init__.py", + "src/transformers/models/minicpm3/configuration_minicpm3.py", + "src/transformers/models/minicpm3/modeling_minicpm3.py", + "src/transformers/models/minicpm3/modular_minicpm3.py", + "tests/models/minicpm3/__init__.py", + "tests/models/minicpm3/test_modeling_minicpm3.py" + ], + "shared_file_areas": [ + { + "filename": "docs/source/en/_toctree.yml", + "left_ranges": [ + [ + 713, + 720 + ] + ], + "right_ranges": [ + [ + 730, + 737 + ] + ] + }, + { + "filename": "docs/source/en/model_doc/minicpm3.md", + "left_ranges": [ + [ + 1, + 67 + ] + ], + "right_ranges": [ + [ + 1, + 45 + ] + ] + }, + { + "filename": "src/transformers/models/auto/modeling_auto.py", + "left_ranges": [ + [ + 285, + 291 + ], + [ + 700, + 706 + ], + [ + 1298, + 1304 + ] + ], + "right_ranges": [ + [ + 720, + 726 + ], + [ + 1331, + 1337 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/__init__.py", + "left_ranges": [ + [ + 1, + 29 + ] + ], + "right_ranges": [ + [ + 1, + 29 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/configuration_minicpm3.py", + "left_ranges": [ + [ + 1, + 141 + ] + ], + "right_ranges": [ + [ + 1, + 126 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 544 + ] + ], + "right_ranges": [ + [ + 1, + 522 + ] + ] + }, + { + "filename": "src/transformers/models/minicpm3/modular_minicpm3.py", + "left_ranges": [ + [ + 1, + 444 + ] + ], + "right_ranges": [ + [ + 1, + 342 + ] + ] + }, + { + "filename": "tests/models/minicpm3/test_modeling_minicpm3.py", + "left_ranges": [ + [ + 1, + 109 + ] + ], + "right_ranges": [ + [ + 1, + 136 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45647-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45647.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 45647, + "canonical_pr_number": 45738, + "issue_numbers": [ + 45647 + ], + "pr_numbers": [ + 45737, + 45738 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 45737, + "right_pr_number": 45738, + "code_similarity": 0.526, + "size_similarity": 0.11, + "file_overlap": 0.5, + "area_overlap": 1.0, + "patch_similarity": 0.024, + "shared_filenames": [ + "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/musicgen_melody/modeling_musicgen_melody.py", + "left_ranges": [ + [ + 491, + 497 + ] + ], + "right_ranges": [ + [ + 491, + 497 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-45987-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #45987.", + "status": "open", + "confidence": 0.75, + "canonical_issue_number": 45987, + "canonical_pr_number": 46173, + "issue_numbers": [ + 45987 + ], + "pr_numbers": [ + 46173, + 46474 + ], + "evidence_types": [ + "closing_reference", + "shared_issue_target" + ], + "pr_comparisons": [ + { + "left_pr_number": 46173, + "right_pr_number": 46474, + "code_similarity": 1.0, + "size_similarity": 1.0, + "file_overlap": 1.0, + "area_overlap": 1.0, + "patch_similarity": 1.0, + "shared_filenames": [ + "src/transformers/cache_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/cache_utils.py", + "left_ranges": [ + [ + 69, + 75 + ], + [ + 375, + 381 + ] + ], + "right_ranges": [ + [ + 69, + 75 + ], + [ + 375, + 381 + ] + ] + } + ] + } + ] + }, + { + "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": 0.494, + "size_similarity": 0.619, + "file_overlap": 0.5, + "area_overlap": 0.434, + "patch_similarity": 0.451, + "shared_filenames": [ + "src/transformers/testing_utils.py", + "tests/utils/test_testing_utils.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/testing_utils.py", + "left_ranges": [ + [ + 3595, + 3602 + ], + [ + 3819, + 3853 + ] + ], + "right_ranges": [ + [ + 3525, + 3558 + ], + [ + 3782, + 3788 + ] + ] + }, + { + "filename": "tests/utils/test_testing_utils.py", + "left_ranges": [ + [ + 1, + 86 + ] + ], + "right_ranges": [ + [ + 1, + 114 + ] + ] + } + ] + } + ] + }, + { + "cluster_id": "cluster-46291-3", + "summary": "Cluster of 1 issues and 2 PRs centered on issue #46291.", + "status": "open", + "confidence": 0.55, + "canonical_issue_number": 46291, + "canonical_pr_number": 46293, + "issue_numbers": [ + 46291 + ], + "pr_numbers": [ + 46293, + 46478 + ], + "evidence_types": [ + "closing_reference" + ], + "pr_comparisons": [ + { + "left_pr_number": 46293, + "right_pr_number": 46478, + "code_similarity": 0.483, + "size_similarity": 0.325, + "file_overlap": 0.333, + "area_overlap": 0.862, + "patch_similarity": 0.111, + "shared_filenames": [ + "src/transformers/integrations/hub_kernels.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/integrations/hub_kernels.py", + "left_ranges": [ + [ + 89, + 180 + ], + [ + 233, + 247 + ] + ], + "right_ranges": [ + [ + 89, + 181 + ], + [ + 234, + 248 + ], + [ + 263, + 269 + ], + [ + 360, + 366 + ] + ] + } + ] + } + ] + }, + { + "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-44018-2", + "summary": "Cluster of 2 related pull requests linked by soft_similarity.", + "status": "open", + "confidence": 0.5, + "canonical_issue_number": null, + "canonical_pr_number": 44068, + "issue_numbers": [], + "pr_numbers": [ + 44018, + 44068 + ], + "evidence_types": [ + "soft_similarity" + ], + "pr_comparisons": [ + { + "left_pr_number": 44018, + "right_pr_number": 44068, + "code_similarity": 0.766, + "size_similarity": 0.939, + "file_overlap": 1.0, + "area_overlap": 0.425, + "patch_similarity": 0.866, + "shared_filenames": [ + "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "tests/models/gpt_neo/test_modeling_gpt_neo.py" + ], + "shared_file_areas": [ + { + "filename": "src/transformers/models/gpt_neo/modeling_gpt_neo.py", + "left_ranges": [ + [ + 26, + 46 + ], + [ + 138, + 143 + ], + [ + 182, + 187 + ], + [ + 283, + 295 + ], + [ + 327, + 341 + ], + [ + 348, + 354 + ], + [ + 360, + 369 + ], + [ + 399, + 406 + ], + [ + 411, + 419 + ], + [ + 428, + 433 + ], + [ + 472, + 492 + ], + [ + 507, + 513 + ], + [ + 519, + 528 + ], + [ + 541, + 559 + ], + [ + 562, + 567 + ], + [ + 595, + 601 + ], + [ + 607, + 614 + ], + [ + 627, + 643 + ], + [ + 685, + 690 + ], + [ + 708, + 714 + ], + [ + 720, + 727 + ], + [ + 740, + 757 + ], + [ + 761, + 766 + ], + [ + 780, + 786 + ], + [ + 791, + 798 + ], + [ + 807, + 822 + ], + [ + 840, + 845 + ] + ], + "right_ranges": [ + [ + 26, + 31 + ], + [ + 34, + 44 + ], + [ + 136, + 141 + ], + [ + 180, + 185 + ], + [ + 281, + 293 + ], + [ + 325, + 339 + ], + [ + 346, + 352 + ], + [ + 358, + 367 + ], + [ + 397, + 404 + ], + [ + 409, + 417 + ], + [ + 426, + 434 + ], + [ + 463, + 483 + ], + [ + 498, + 504 + ], + [ + 510, + 519 + ], + [ + 532, + 550 + ], + [ + 553, + 564 + ], + [ + 586, + 592 + ], + [ + 598, + 605 + ], + [ + 618, + 634 + ], + [ + 676, + 688 + ], + [ + 699, + 705 + ], + [ + 711, + 718 + ], + [ + 731, + 748 + ], + [ + 752, + 762 + ], + [ + 771, + 777 + ], + [ + 782, + 789 + ], + [ + 798, + 813 + ], + [ + 831, + 836 + ] + ] + }, + { + "filename": "tests/models/gpt_neo/test_modeling_gpt_neo.py", + "left_ranges": [ + [ + 458, + 464 + ] + ], + "right_ranges": [ + [ + 458, + 464 + ] + ] + } + ] + } + ] + } + ], + "duplicate_issues": [], + "duplicate_prs": [ + { + "cluster_id": "cluster-41115-3", + "canonical_pr_number": 45613, + "duplicate_pr_numbers": [ + 41116 + ], + "target_issue_number": 41115, + "reason": "PRs in cluster-41115-3 are treated as duplicates because they converge on issue #41115 with closing_reference evidence." + }, + { + "cluster_id": "cluster-41211-3", + "canonical_pr_number": 44339, + "duplicate_pr_numbers": [ + 41356 + ], + "target_issue_number": 41211, + "reason": "PRs in cluster-41211-3 are treated as duplicates because they converge on issue #41211 with closing_reference evidence." + }, + { + "cluster_id": "cluster-43240-3", + "canonical_pr_number": 43251, + "duplicate_pr_numbers": [ + 43254 + ], + "target_issue_number": 43240, + "reason": "PRs in cluster-43240-3 are treated as duplicates because they converge on issue #43240 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43698-3", + "canonical_pr_number": 43779, + "duplicate_pr_numbers": [ + 43816 + ], + "target_issue_number": 43698, + "reason": "PRs in cluster-43698-3 are treated as duplicates because they converge on issue #43698 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-43979-11", + "canonical_pr_number": 44007, + "duplicate_pr_numbers": [ + 43996, + 44013, + 44044, + 44066, + 44072, + 44085, + 44129, + 44154, + 44722 + ], + "target_issue_number": 43979, + "reason": "PRs in cluster-43979-11 are treated as duplicates because they converge on issue #43979 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-44018-2", + "canonical_pr_number": 44068, + "duplicate_pr_numbers": [ + 44018 + ], + "target_issue_number": null, + "reason": "PRs in cluster-44018-2 are treated as duplicates because they share soft_similarity evidence." + }, + { + "cluster_id": "cluster-45081-3", + "canonical_pr_number": 45317, + "duplicate_pr_numbers": [ + 45086 + ], + "target_issue_number": 45081, + "reason": "PRs in cluster-45081-3 are treated as duplicates because they converge on issue #45081 with closing_reference, shared_issue_target, soft_similarity evidence." + }, + { + "cluster_id": "cluster-45363-3", + "canonical_pr_number": 46339, + "duplicate_pr_numbers": [ + 45363, + 45666 + ], + "target_issue_number": null, + "reason": "PRs in cluster-45363-3 are treated as duplicates because they share 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-45987-3", + "canonical_pr_number": 46173, + "duplicate_pr_numbers": [ + 46474 + ], + "target_issue_number": 45987, + "reason": "PRs in cluster-45987-3 are treated as duplicates because they converge on issue #45987 with closing_reference, shared_issue_target evidence." + }, + { + "cluster_id": "cluster-46291-3", + "canonical_pr_number": 46293, + "duplicate_pr_numbers": [ + 46478 + ], + "target_issue_number": 46291, + "reason": "PRs in cluster-46291-3 are treated as duplicates because they converge on issue #46291 with closing_reference evidence." + } + ], + "best_issue": { + "cluster_id": "cluster-41211-3", + "issue_number": 41211, + "reason": "Issue #41211 is the strongest global issue candidate because it is open, belongs to a cluster with 3 artifacts, and carries 6 discussion comments plus 6 inbound references.", + "score": 148.0 + }, + "best_pr": { + "cluster_id": "cluster-43979-11", + "pr_number": 44007, + "reason": "PR #44007 is the strongest global PR candidate because it is open, belongs to a cluster with 11 artifacts, links to 1 issues, and carries 2 review/discussion events.", + "score": 238.886 + } +} diff --git a/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json new file mode 100644 index 0000000000000000000000000000000000000000..7a6d68ada704220d50e07725073c84b4cfad2445 --- /dev/null +++ b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json @@ -0,0 +1,22419 @@ +{ + "schema_version": "1.0", + "repo": "huggingface/transformers", + "snapshot_id": "20260608T180018Z", + "generated_at": "2026-06-08T18:11:33Z", + "model": "gpt-5.4-mini?service_tier=flex", + "reviews": [ + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:24643", + "issue:30277", + "issue:34634", + "issue:34928", + "issue:35545", + "issue:36010", + "issue:36331", + "issue:39290", + "issue:41720", + "issue:41762", + "issue:42915", + "issue:43039", + "issue:43296", + "issue:43425", + "issue:43541", + "issue:43596", + "issue:43638", + "issue:43716", + "issue:43827", + "issue:43828", + "issue:43866", + "issue:43975", + "issue:44292", + "issue:44322", + "issue:44457", + "issue:44509", + "issue:44512", + "issue:44560", + "issue:44805", + "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:46304", + "issue:46402", + "issue:46468" + ], + "soft_pairs": [ + "issue:44322|issue:45464", + "issue:45237|issue:45685", + "issue:43716|issue:45237", + "issue:43828|issue:45237", + "issue:45237|issue:46468", + "issue:45237|issue:45507", + "issue:30277|issue:35545", + "issue:35545|issue:42915", + "issue:35545|issue:41720", + "issue:45901|issue:46015", + "issue:44509|issue:44512", + "issue:35545|issue:36010", + "issue:44805|issue:45823", + "issue:44918|issue:46165", + "issue:43638|issue:44805", + "issue:43638|issue:46165", + "issue:43596|issue:44805", + "issue:44805|issue:46294", + "issue:44805|issue:45161", + "issue:46165|issue:46468", + "issue:44560|issue:46165", + "issue:43596|issue:46165", + "issue:34634|issue:35545", + "issue:30277|issue:43638", + "issue:43596|issue:43638", + "issue:43039|issue:45161", + "issue:43975|issue:44457", + "issue:44292|issue:44457", + "issue:24643|issue:30277", + "issue:34928|issue:44805", + "issue:45859|issue:45902", + "issue:43638|issue:45161", + "issue:34928|issue:36331", + "issue:43596|issue:45161", + "issue:43716|issue:45507", + "issue:45874|issue:46294", + "issue:44292|issue:45237", + "issue:39290|issue:45081", + "issue:45823|issue:45874", + "issue:43425|issue:44292", + "issue:43828|issue:45161", + "issue:34928|issue:41720", + "issue:43828|issue:44292", + "issue:43541|issue:43716", + "issue:43716|issue:46468", + "issue:43716|issue:45685", + "issue:44918|issue:45161", + "issue:44292|issue:44918", + "issue:43716|issue:44292", + "issue:45507|issue:45685", + "issue:43039|issue:44918", + "issue:41762|issue:43638", + "issue:43827|issue:44512", + "issue:43828|issue:45685", + "issue:44918|issue:45685", + "issue:43039|issue:43828", + "issue:30277|issue:36331", + "issue:34928|issue:36010", + "issue:45237|issue:46015", + "issue:43039|issue:44805", + "issue:43866|issue:46402", + "issue:45800|issue:46304", + "issue:43296|issue:46304", + "issue:41762|issue:43596", + "issue:39290|issue:41720" + ], + "prepared_review_unit_hash": "ee8bb41fda9a60efc1f33b4f5d6dab030c88af11c8c89b9b9f6e0b4eda4ca225", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 65, + "serialized_chars": 47171, + "estimated_input_tokens": 11793, + "estimated_eval_tokens": 23842 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are clearly unrelated despite superficial similarity. The only strong duplicate pair is the ZeRO-3/BERT IndexError, and there is one plausible Qwen3_5Moe rotary-embedding device/materialization duplicate pair.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43596 is the cleanest representative of the confirmed ZeRO-3/BERT IndexError duplicate; issue:43638 is the same failure with a more specific phrasing.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43596 is the best issue to keep: it is well-scoped, generic enough to cover the duplicate, and matches the clearest duplicate partner.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44322", + "right": "issue:45464", + "accept": false, + "reason": "Different models and failure modes; not the same bug." + }, + { + "left": "issue:45237", + "right": "issue:45685", + "accept": false, + "reason": "Different backend/model issue; only broad MoE/hardware similarity." + }, + { + "left": "issue:43716", + "right": "issue:45237", + "accept": false, + "reason": "Different model-specific runtime errors." + }, + { + "left": "issue:43828", + "right": "issue:45237", + "accept": false, + "reason": "Both are dtype/backend problems, but not the same code path." + }, + { + "left": "issue:45237", + "right": "issue:46468", + "accept": false, + "reason": "Different model/example failures." + }, + { + "left": "issue:45237", + "right": "issue:45507", + "accept": false, + "reason": "Different model families and symptoms." + }, + { + "left": "issue:30277", + "right": "issue:35545", + "accept": false, + "reason": "DeepSpeed collective mismatch vs ONNX export error are unrelated." + }, + { + "left": "issue:35545", + "right": "issue:42915", + "accept": false, + "reason": "Different model/export issues." + }, + { + "left": "issue:35545", + "right": "issue:41720", + "accept": false, + "reason": "ONNX export vs device-map cuda assert; unrelated." + }, + { + "left": "issue:45901", + "right": "issue:46015", + "accept": false, + "reason": "Task crash vs documentation issue; not a duplicate." + }, + { + "left": "issue:44509", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs cleanup, but they target different removed references." + }, + { + "left": "issue:35545", + "right": "issue:36010", + "accept": false, + "reason": "Different import/export failures." + }, + { + "left": "issue:44805", + "right": "issue:45823", + "accept": false, + "reason": "Generic mask-shape error vs Gemma4 device mismatch are not the same bug." + }, + { + "left": "issue:44918", + "right": "issue:46165", + "accept": false, + "reason": "Different training/resume problems." + }, + { + "left": "issue:43638", + "right": "issue:44805", + "accept": false, + "reason": "Different root causes and model paths." + }, + { + "left": "issue:43638", + "right": "issue:46165", + "accept": false, + "reason": "ZeRO-3/BERT init error vs trust-remote-code resume issue." + }, + { + "left": "issue:43596", + "right": "issue:44805", + "accept": false, + "reason": "Different IndexError shapes and code paths." + }, + { + "left": "issue:44805", + "right": "issue:46294", + "accept": false, + "reason": "Different multimodal input validation failures." + }, + { + "left": "issue:44805", + "right": "issue:45161", + "accept": false, + "reason": "Mask-shape error vs TP/MoE failure." + }, + { + "left": "issue:46165", + "right": "issue:46468", + "accept": false, + "reason": "Different model/training issues." + }, + { + "left": "issue:44560", + "right": "issue:46165", + "accept": false, + "reason": "Video embedding error vs training resume issue." + }, + { + "left": "issue:43596", + "right": "issue:46165", + "accept": false, + "reason": "Different problems." + }, + { + "left": "issue:34634", + "right": "issue:35545", + "accept": false, + "reason": "Processor preset issue vs ONNX export error." + }, + { + "left": "issue:30277", + "right": "issue:43638", + "accept": false, + "reason": "Different ZeRO-3 failures on different models/code paths." + }, + { + "left": "issue:43596", + "right": "issue:43638", + "accept": true, + "reason": "Same ZeRO-3/BERT IndexError at load/init; only phrasing differs." + }, + { + "left": "issue:43039", + "right": "issue:45161", + "accept": false, + "reason": "Different features and failure modes." + }, + { + "left": "issue:43975", + "right": "issue:44457", + "accept": false, + "reason": "Detokenization bug vs LoRA save/load mismatch." + }, + { + "left": "issue:44292", + "right": "issue:44457", + "accept": false, + "reason": "Different model/runtime issue vs LoRA merge mismatch." + }, + { + "left": "issue:24643", + "right": "issue:30277", + "accept": false, + "reason": "Different DeepSpeed errors." + }, + { + "left": "issue:34928", + "right": "issue:44805", + "accept": false, + "reason": "Activation-checkpoint/FSDP recompute mismatch vs mask-shape IndexError." + }, + { + "left": "issue:45859", + "right": "issue:45902", + "accept": true, + "reason": "Same Qwen3_5MoeTextRotaryEmbedding device/materialization bug; one is the CPU-offload symptom, the other the uninitialized inv_freq symptom." + }, + { + "left": "issue:44292", + "right": "issue:45237", + "accept": false, + "reason": "Different model/runtime failures." + }, + { + "left": "issue:39290", + "right": "issue:45081", + "accept": false, + "reason": "Different model/version regressions." + }, + { + "left": "issue:45823", + "right": "issue:45874", + "accept": false, + "reason": "Different Gemma4 issues." + }, + { + "left": "issue:43425", + "right": "issue:44292", + "accept": false, + "reason": "Torch compatibility issue vs model runtime error." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Dtype mismatch vs MPS histogram int error are unrelated." + }, + { + "left": "issue:34928", + "right": "issue:41720", + "accept": false, + "reason": "Different distributed-training and device-map failures." + }, + { + "left": "issue:43828", + "right": "issue:44292", + "accept": false, + "reason": "Different model-specific failures." + }, + { + "left": "issue:43541", + "right": "issue:43716", + "accept": false, + "reason": "Different model/tracing vs image-preprocessor dtype issues." + }, + { + "left": "issue:43716", + "right": "issue:46468", + "accept": false, + "reason": "Different model-specific bugs." + }, + { + "left": "issue:43716", + "right": "issue:45685", + "accept": false, + "reason": "Different device/dtype backends." + }, + { + "left": "issue:44918", + "right": "issue:45161", + "accept": false, + "reason": "Different training feature failures." + }, + { + "left": "issue:44292", + "right": "issue:44918", + "accept": false, + "reason": "Different model/runtime bugs." + }, + { + "left": "issue:43716", + "right": "issue:44292", + "accept": false, + "reason": "Different issues with only loose multimodal overlap." + }, + { + "left": "issue:45507", + "right": "issue:45685", + "accept": false, + "reason": "Different methods/backends and symptoms." + }, + { + "left": "issue:43039", + "right": "issue:44918", + "accept": false, + "reason": "Liger kernel cross-entropy issue vs SFT unpacking error." + }, + { + "left": "issue:41762", + "right": "issue:43638", + "accept": false, + "reason": "Both involve ZeRO-3 IndexError, but on different models and likely different code paths." + }, + { + "left": "issue:43827", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs regressions, but they cover different removed references." + }, + { + "left": "issue:43828", + "right": "issue:45685", + "accept": false, + "reason": "Dtype mismatch vs MPS int histogram error are unrelated." + }, + { + "left": "issue:44918", + "right": "issue:45685", + "accept": false, + "reason": "Different training/model failures." + }, + { + "left": "issue:43039", + "right": "issue:43828", + "accept": false, + "reason": "Different bugs." + }, + { + "left": "issue:30277", + "right": "issue:36331", + "accept": false, + "reason": "DeepSpeed collective mismatch vs trainer API signature break." + }, + { + "left": "issue:34928", + "right": "issue:36010", + "accept": false, + "reason": "Different subsystems and failure modes." + }, + { + "left": "issue:45237", + "right": "issue:46015", + "accept": false, + "reason": "Model/runtime failure vs documentation issue." + }, + { + "left": "issue:43039", + "right": "issue:44805", + "accept": false, + "reason": "Different code paths and symptoms." + }, + { + "left": "issue:43866", + "right": "issue:46402", + "accept": false, + "reason": "Checkpoint corruption vs missing weight loading." + }, + { + "left": "issue:45800", + "right": "issue:46304", + "accept": false, + "reason": "Different compatibility regressions." + }, + { + "left": "issue:43296", + "right": "issue:46304", + "accept": false, + "reason": "Different release incompatibilities." + }, + { + "left": "issue:41762", + "right": "issue:43596", + "accept": false, + "reason": "Same broad ZeRO-3 error shape, but different models and contexts." + }, + { + "left": "issue:39290", + "right": "issue:41720", + "accept": false, + "reason": "Different model/version-specific failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative on the soft edges. The two accepted duplicates are consistent with the titles/body signals: 43596\u219443638 and 45859\u219445902. The rejections are also appropriately cautious and do not overclaim duplication from superficial similarity." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 2, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36010", + "issue:39290", + "issue:41093", + "issue:41762", + "issue:43039", + "issue:43296", + "issue:43366", + "issue:43404", + "issue:43425", + "issue:43531", + "issue:43541", + "issue:43572", + "issue:43596", + "issue:43638", + "issue:43792", + "issue:43828", + "issue:43854", + "issue:43866", + "issue:43872", + "issue:43901", + "issue:43975", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44512", + "issue:44560", + "issue:44589", + "issue:44805", + "issue:44841", + "issue:44863", + "issue:44918", + "issue:45005", + "issue:45020", + "issue:45070", + "issue:45084", + "issue:45161", + "issue:45237", + "issue:45362", + "issue:45464", + "issue:45685", + "issue:45800", + "issue:45812", + "issue:45823", + "issue:45901", + "issue:45920", + "issue:46015", + "issue:46294", + "issue:46304" + ], + "soft_pairs": [ + "issue:44589|issue:46304", + "issue:45823|issue:46294", + "issue:43901|issue:44512", + "issue:45237|issue:45901", + "issue:44841|issue:45823", + "issue:43039|issue:44560", + "issue:43975|issue:46015", + "issue:44863|issue:46015", + "issue:45020|issue:46304", + "issue:45812|issue:45920", + "issue:43866|issue:44863", + "issue:44292|issue:46015", + "issue:43872|issue:46304", + "issue:39290|issue:43531", + "issue:41093|issue:43638", + "issue:43792|issue:45901", + "issue:43854|issue:43866", + "issue:43425|issue:45070", + "issue:43596|issue:45800", + "issue:43975|issue:45901", + "issue:43792|issue:45070", + "issue:45362|issue:45464", + "issue:44292|issue:45901", + "issue:39290|issue:45823", + "issue:43404|issue:43866", + "issue:45464|issue:45685", + "issue:43541|issue:45685", + "issue:43541|issue:43828", + "issue:44841|issue:45084", + "issue:36010|issue:39290", + "issue:45800|issue:45823", + "issue:41093|issue:41762", + "issue:43296|issue:43366", + "issue:43572|issue:45070", + "issue:44805|issue:44918", + "issue:43541|issue:45161", + "issue:44387|issue:45005", + "issue:43828|issue:45464", + "issue:44291|issue:45070", + "issue:43366|issue:45464", + "issue:43531|issue:45070", + "issue:45020|issue:45800" + ], + "prepared_review_unit_hash": "d1c1b356f5ad4477df0f519a94c4e725c2085e8565289da3df458ca723d25a6e", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 42, + "serialized_chars": 40100, + "estimated_input_tokens": 10025, + "estimated_eval_tokens": 20306 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is dominated by one clear umbrella issue: the v5 AutoTokenizer wrong-token-ID regression. Most other pairs are unrelated single-model, docs, or compatibility bugs. The DeepSpeed ZeRO-3/Bert index-error pair is now treated conservatively as non-duplicate because the shared symptom is too generic to prove one root cause.", + "confidence": 0.83, + "canonical_issue_reason": "issue:45920 is the broadest and most representative tokenizer regression report: it spans multiple affected model families, is still open, and has inbound references, making it the best umbrella issue for the duplicate cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the best cluster representative because it captures the widest scope of the tokenizer-ID regression across several models, not just one family.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44589", + "right": "issue:46304", + "accept": false, + "reason": "Different failures: a missing storage class vs a torch/distributed compatibility regression." + }, + { + "left": "issue:45823", + "right": "issue:46294", + "accept": false, + "reason": "Different model paths and symptoms: Gemma4 device mismatch vs Qwen3.5 video/chat-template token mismatch." + }, + { + "left": "issue:43901", + "right": "issue:44512", + "accept": false, + "reason": "Both are docs issues, but they concern unrelated documentation mismatches." + }, + { + "left": "issue:45237", + "right": "issue:45901", + "accept": false, + "reason": "Different subsystems and failure modes: AMD GPU support vs table QA task crash." + }, + { + "left": "issue:44841", + "right": "issue:45823", + "accept": false, + "reason": "Different models and bugs: Voxtral processor failure vs Gemma4 device placement issue." + }, + { + "left": "issue:43039", + "right": "issue:44560", + "accept": false, + "reason": "Liger kernel cross-entropy routing and Qwen3-VL video StopIteration are unrelated." + }, + { + "left": "issue:43975", + "right": "issue:46015", + "accept": false, + "reason": "Model detokenization bug vs outdated RAG docs; not the same underlying issue." + }, + { + "left": "issue:44863", + "right": "issue:46015", + "accept": false, + "reason": "NemotronH checkpoint-loading failure is unrelated to stale RAG examples." + }, + { + "left": "issue:45020", + "right": "issue:46304", + "accept": false, + "reason": "Remote_code breakage and USE_DISTRIBUTED=0 support are different compatibility problems." + }, + { + "left": "issue:45812", + "right": "issue:45920", + "accept": true, + "reason": "Same underlying tokenizer regression: both report wrong token IDs from AutoTokenizer in v5 across multiple models." + }, + { + "left": "issue:43866", + "right": "issue:44863", + "accept": false, + "reason": "Ovis2 checkpoint corruption and NemotronH loading failure are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:46015", + "accept": false, + "reason": "A model loading error and a docs issue do not share the same bug." + }, + { + "left": "issue:43872", + "right": "issue:46304", + "accept": false, + "reason": "bitsandbytes constructor incompatibility is unrelated to distributed support changes." + }, + { + "left": "issue:39290", + "right": "issue:43531", + "accept": false, + "reason": "Both mention sliding_window, but they affect different model families and look like separate bugs." + }, + { + "left": "issue:41093", + "right": "issue:43638", + "accept": false, + "reason": "Both are DeepSpeed ZeRO-3/Bert index errors, but the shared out-of-bounds symptom is too generic to confirm one concrete root cause." + }, + { + "left": "issue:43792", + "right": "issue:45901", + "accept": false, + "reason": "Whisper runtime failure and table QA crash are unrelated." + }, + { + "left": "issue:43854", + "right": "issue:43866", + "accept": false, + "reason": "Different model-loading problems with no shared underlying bug." + }, + { + "left": "issue:43425", + "right": "issue:45070", + "accept": false, + "reason": "Torch version incompatibility and pydantic config breakage are different regression types." + }, + { + "left": "issue:43596", + "right": "issue:45800", + "accept": false, + "reason": "ZeRO-3/Bert index error is unrelated to torch 2.4.1 vs transformers 5.8.0 compatibility." + }, + { + "left": "issue:43975", + "right": "issue:45901", + "accept": false, + "reason": "Detokenization bug and table QA crash are unrelated." + }, + { + "left": "issue:43792", + "right": "issue:45070", + "accept": false, + "reason": "Whisper loading/runtime failure is unrelated to PretrainedConfig pydantic-model changes." + }, + { + "left": "issue:45362", + "right": "issue:45464", + "accept": false, + "reason": "Same model family, but different surfaces and insufficient evidence of one concrete shared bug." + }, + { + "left": "issue:45685", + "right": "issue:45901", + "accept": false, + "reason": "MPS MoE histogram error and table QA crash are unrelated." + }, + { + "left": "issue:45800", + "right": "issue:45823", + "accept": false, + "reason": "Torch version incompatibility and Gemma4 device mismatch are separate issues." + }, + { + "left": "issue:41093", + "right": "issue:41762", + "accept": false, + "reason": "Both are ZeRO-3/Bert-related index errors, but the concrete failures differ enough that they do not clearly collapse to one bug." + }, + { + "left": "issue:43296", + "right": "issue:43366", + "accept": false, + "reason": "PaddleOCR-VL loading in vLLM and GGUF gpt-oss support are different issues." + }, + { + "left": "issue:43572", + "right": "issue:45070", + "accept": false, + "reason": "Config-field absence in StableLm and a Pydantic PretrainedConfig break are related only at a very broad level." + }, + { + "left": "issue:44805", + "right": "issue:44918", + "accept": false, + "reason": "They both fail with index-related errors, but one is a generic mask-shape mismatch and the other is a trainer embedding unpacking failure." + }, + { + "left": "issue:43541", + "right": "issue:45161", + "accept": false, + "reason": "Both involve MoE models, but the grouped_mm tracing error and TP-only GPT-OSS issue are not the same bug." + }, + { + "left": "issue:44387", + "right": "issue:45005", + "accept": false, + "reason": "Int4 OOM and tied-weights translation issues are unrelated regressions." + }, + { + "left": "issue:43828", + "right": "issue:45464", + "accept": false, + "reason": "Autocast dtype mismatch and Qwen3.5 streaming API failure are different failures." + }, + { + "left": "issue:44291", + "right": "issue:45070", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized TypeError is unrelated to the pydantic PretrainedConfig regression." + }, + { + "left": "issue:43366", + "right": "issue:45464", + "accept": false, + "reason": "GGUF gpt-oss support and Qwen3.5 streaming chat failure are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:45070", + "accept": false, + "reason": "Sliding-window model config issue and PretrainedConfig pydantic break are different bugs." + }, + { + "left": "issue:45020", + "right": "issue:45800", + "accept": false, + "reason": "Remote_code regression and torch/transformers version incompatibility are separate compatibility problems." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall representative issue choice is grounded, especially issue:45920 as the broadest tokenizer-ID regression report. However, the soft-edge acceptance of issue:41093 \u2194 issue:43638 is too strong for the packet: both are DeepSpeed/ZeRO-3/Bert-related index errors, but the titles only show broadly similar failure types, not a clearly shared root cause. That makes the verdict insufficiently conservative. The other soft-edge rejections look fine." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 3, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36331", + "issue:39186", + "issue:39692", + "issue:40990", + "issue:41093", + "issue:41762", + "issue:42915", + "issue:43054", + "issue:43404", + "issue:43541", + "issue:43596", + "issue:43644", + "issue:43645", + "issue:43716", + "issue:43828", + "issue:43872", + "issue:43873", + "issue:43950", + "issue:43975", + "issue:43994", + "issue:44292", + "issue:44360", + "issue:44479", + "issue:44492", + "issue:44512", + "issue:44530", + "issue:44534", + "issue:44589", + "issue:44863", + "issue:44918", + "issue:44928", + "issue:44960", + "issue:45005", + "issue:45183", + "issue:45399", + "issue:45507", + "issue:45542", + "issue:45632", + "issue:45698", + "issue:45759", + "issue:45800", + "issue:45902", + "issue:45920", + "issue:46294", + "issue:46463", + "issue:46489", + "issue:46491" + ], + "soft_pairs": [ + "issue:43716|issue:46294", + "issue:43541|issue:44589", + "issue:44479|issue:46294", + "issue:41093|issue:43596", + "issue:43054|issue:43994", + "issue:45759|issue:45902", + "issue:45542|issue:45800", + "issue:43716|issue:44918", + "issue:39186|issue:44918", + "issue:45632|issue:45698", + "issue:43645|issue:45399", + "issue:43404|issue:45005", + "issue:40990|issue:44960", + "issue:42915|issue:45399", + "issue:44530|issue:45005", + "issue:43873|issue:45005", + "issue:44360|issue:44512", + "issue:45183|issue:46463", + "issue:46489|issue:46491", + "issue:44863|issue:45005", + "issue:44918|issue:45507", + "issue:44492|issue:44512", + "issue:43975|issue:44292", + "issue:44292|issue:45507", + "issue:43950|issue:44534", + "issue:39692|issue:43054", + "issue:43644|issue:43950", + "issue:36331|issue:41762", + "issue:43828|issue:44928", + "issue:43872|issue:44589", + "issue:45920|issue:46489", + "issue:43950|issue:45902" + ], + "prepared_review_unit_hash": "deff84affc955a481cfb30bd309559626bc13c7a5418af49e02e128eb1f31b46", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36859, + "estimated_input_tokens": 9215, + "estimated_eval_tokens": 18686 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a mixed set of unrelated Transformers bug reports. The only clear duplicate cluster here is the v5 non-persistent-buffer corruption regression, where 43644 and 44534 match 43950 closely.", + "confidence": 0.88, + "canonical_issue_reason": "43950 is the clearest representative of the non-persistent-buffer regression: it states the v5.x corruption bug directly and matches the other clear duplicate reports.", + "canonical_pr_reason": null, + "best_issue_reason": "43950 is the best single issue to anchor this cluster because it is the most specific and directly describes the shared bug seen in the duplicate pair.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43716", + "right": "issue:46294", + "accept": false, + "reason": "Different concrete failures: image/text dtype mismatch vs video chat-template token/feature mismatch." + }, + { + "left": "issue:43541", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated runtime paths: torch dynamo grouped_mm tracing error vs float8 storage lookup failure." + }, + { + "left": "issue:44479", + "right": "issue:46294", + "accept": false, + "reason": "Both involve Qwen video input, but the reported symptoms and code paths are different enough that this is not a safe duplicate." + }, + { + "left": "issue:41093", + "right": "issue:43596", + "accept": false, + "reason": "Both are index/shape errors around init, but they affect different models and failure points." + }, + { + "left": "issue:43054", + "right": "issue:43994", + "accept": false, + "reason": "Same model family, but one is a text-embedding quality complaint and the other is a broader pipeline/AutoModel output issue." + }, + { + "left": "issue:45759", + "right": "issue:45902", + "accept": false, + "reason": "Different bugs: composite config unwrapping vs uninitialized rotary embedding memory after materialization." + }, + { + "left": "issue:45542", + "right": "issue:45800", + "accept": false, + "reason": "Completely unrelated: TensorFlow backend detection vs torch/transformers version incompatibility." + }, + { + "left": "issue:43716", + "right": "issue:44918", + "accept": false, + "reason": "Different models and different multimodal/trainer failure modes." + }, + { + "left": "issue:39186", + "right": "issue:44918", + "accept": false, + "reason": "FSDP weight-shape error is unrelated to TRL embedding unpacking." + }, + { + "left": "issue:45632", + "right": "issue:45698", + "accept": false, + "reason": "Both concern custom-module loading, but one is a leaf-dir cache collision and the other is wrong module selection after save/load; not the same bug." + }, + { + "left": "issue:43645", + "right": "issue:45399", + "accept": false, + "reason": "Jupyter custom model initialization issue is unrelated to flash-attn fallback gating." + }, + { + "left": "issue:43404", + "right": "issue:45005", + "accept": false, + "reason": "Specific Mistral3 weight-tying bug vs a broader translation-model tied-weights regression; too different to merge." + }, + { + "left": "issue:40990", + "right": "issue:44960", + "accept": false, + "reason": "No meaningful overlap beyond being model-quality reports." + }, + { + "left": "issue:42915", + "right": "issue:45399", + "accept": false, + "reason": "FineGrainedFP8Config failure is unrelated to flash-attn fallback checks." + }, + { + "left": "issue:44530", + "right": "issue:45005", + "accept": false, + "reason": "PagedAttentionCache group-type crash is unrelated to tied-weight regressions." + }, + { + "left": "issue:43873", + "right": "issue:45005", + "accept": false, + "reason": "Quantization/offloading behavior is unrelated to tied-weight handling." + }, + { + "left": "issue:44360", + "right": "issue:44512", + "accept": false, + "reason": "An algorithmic DSA/indexer complaint is unrelated to a docs typo." + }, + { + "left": "issue:45183", + "right": "issue:46463", + "accept": false, + "reason": "Related validation area, but not the same concrete bug or code path." + }, + { + "left": "issue:46489", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions for different model families." + }, + { + "left": "issue:44863", + "right": "issue:45005", + "accept": false, + "reason": "Checkpoint-loading failure is unrelated to tied-weights regression." + }, + { + "left": "issue:44918", + "right": "issue:45507", + "accept": false, + "reason": "Different model families and unrelated failure modes." + }, + { + "left": "issue:44492", + "right": "issue:44512", + "accept": false, + "reason": "Unrelated typo vs v5 docs command removal." + }, + { + "left": "issue:43975", + "right": "issue:44292", + "accept": false, + "reason": "Both are tokenizer/model issues, but they concern different models and different symptoms." + }, + { + "left": "issue:44292", + "right": "issue:45507", + "accept": false, + "reason": "Different model-specific runtime failures." + }, + { + "left": "issue:43950", + "right": "issue:44534", + "accept": true, + "reason": "Same underlying v5 regression where non-persistent buffers get corrupted/fill with junk during load." + }, + { + "left": "issue:39692", + "right": "issue:43054", + "accept": false, + "reason": "Both mention SigLIP2, but one is a docs/example error report and the other is a model-quality regression." + }, + { + "left": "issue:43644", + "right": "issue:43950", + "accept": true, + "reason": "These report the same non-persistent-buffer corruption regression in Transformers v5, phrased slightly differently." + }, + { + "left": "issue:36331", + "right": "issue:41762", + "accept": false, + "reason": "Different errors: custom trainer kwarg mismatch vs ZeRO-3 load-time index error." + }, + { + "left": "issue:43828", + "right": "issue:44928", + "accept": false, + "reason": "Both are mixed-precision/dtype-related, but the models, symptoms, and code paths differ." + }, + { + "left": "issue:43872", + "right": "issue:44589", + "accept": false, + "reason": "Both are quantization/storage compatibility issues, but they are distinct TypeErrors with different root causes." + }, + { + "left": "issue:45920", + "right": "issue:46489", + "accept": false, + "reason": "Broader tokenizer regression vs a specific DeepSeek-Coder v1 tokenization bug; not enough evidence they are the same defect." + }, + { + "left": "issue:43950", + "right": "issue:45902", + "accept": false, + "reason": "Non-persistent-buffer corruption is unrelated to the Qwen3_5Moe rotary-embedding memory bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only clearly supported duplicate set is the non-persistent-buffer regression around 43644/44534/43950, and the other soft-edge verdicts are mostly cautious false negatives rather than overclaims." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 4, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:24643", + "issue:28530", + "issue:38175", + "issue:41553", + "issue:41762", + "issue:42491", + "issue:42915", + "issue:42947", + "issue:43054", + "issue:43404", + "issue:43493", + "issue:43643", + "issue:43742", + "issue:43782", + "issue:43824", + "issue:43828", + "issue:43856", + "issue:43866", + "issue:43872", + "issue:43881", + "issue:43883", + "issue:43950", + "issue:43975", + "issue:44368", + "issue:44387", + "issue:44403", + "issue:44451", + "issue:44488", + "issue:44661", + "issue:44910", + "issue:44960", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45356", + "issue:45399", + "issue:45488", + "issue:45520", + "issue:45663", + "issue:45685", + "issue:45701", + "issue:45715", + "issue:45823", + "issue:45859", + "issue:45902", + "issue:46315", + "issue:46489" + ], + "soft_pairs": [ + "issue:43866|issue:44403", + "issue:38175|issue:43054", + "issue:41553|issue:43883", + "issue:43054|issue:45701", + "issue:44910|issue:45859", + "issue:45663|issue:45902", + "issue:42947|issue:44387", + "issue:41553|issue:43881", + "issue:44403|issue:44451", + "issue:24643|issue:28530", + "issue:43054|issue:43493", + "issue:43975|issue:45520", + "issue:44403|issue:44488", + "issue:42947|issue:43856", + "issue:43643|issue:45020", + "issue:41762|issue:42915", + "issue:45685|issue:45823", + "issue:43881|issue:43883", + "issue:44991|issue:45399", + "issue:43824|issue:43872", + "issue:45020|issue:45823", + "issue:44661|issue:45020", + "issue:43828|issue:45823", + "issue:42491|issue:44368", + "issue:45081|issue:45399", + "issue:45715|issue:46315", + "issue:45488|issue:46489", + "issue:43782|issue:44960", + "issue:45020|issue:45356", + "issue:43742|issue:45399", + "issue:43404|issue:43950" + ], + "prepared_review_unit_hash": "d5000620e0e3b0a15de50f4531aa0b97932ea6adaebaf3473fa39e939da89b95", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36740, + "estimated_input_tokens": 9185, + "estimated_eval_tokens": 18626 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of model-loading/tokenization regressions plus a few unrelated performance/configuration bugs; there is no true duplicate group. If one issue must serve as the anchor, 45020 is the broadest umbrella report.", + "confidence": 0.56, + "canonical_issue_reason": "issue:45020 is the most general and central report: it describes a broad regression where recent transformers versions break remote_code model loading, which overlaps conceptually with several model-specific loading failures.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the best representative issue because it is the broadest umbrella symptom and most likely to subsume related version-regression reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43866", + "right": "issue:44403", + "accept": false, + "reason": "Checkpoint corruption during model load is unrelated to a logging/noise complaint." + }, + { + "left": "issue:38175", + "right": "issue:43054", + "accept": false, + "reason": "Both involve SigLIP/SigLIP2, but one is zero probabilities and the other is embedding quality; different bugs." + }, + { + "left": "issue:41553", + "right": "issue:43883", + "accept": false, + "reason": "Both are loading errors, but Voxtral tokenizer error and Molmo missing attribute are distinct code paths." + }, + { + "left": "issue:43054", + "right": "issue:45701", + "accept": false, + "reason": "Text-embedding quality regression is not the same as a broad tokenization-version change report." + }, + { + "left": "issue:44910", + "right": "issue:45859", + "accept": false, + "reason": "Flash-Attention crash from 3D position_ids is different from CPU-offload incompatibility in rotary embedding." + }, + { + "left": "issue:45663", + "right": "issue:45902", + "accept": false, + "reason": "Both mention Gemma4/Qwen3 MoE internals, but FSDP2 KeyError and uninitialized memory are separate defects." + }, + { + "left": "issue:42947", + "right": "issue:44387", + "accept": false, + "reason": "Gradient checkpointing with PEFT LoRA is unrelated to int4 quantization memory growth." + }, + { + "left": "issue:41553", + "right": "issue:43881", + "accept": false, + "reason": "Voxtral tokenizer loading and glm-4v-9b loading failure are different model-specific issues." + }, + { + "left": "issue:44403", + "right": "issue:44451", + "accept": false, + "reason": "Generic load-time noise is not the same as failing to load a specific model." + }, + { + "left": "issue:24643", + "right": "issue:28530", + "accept": false, + "reason": "DeepSpeed weight-shape runtime error is unrelated to early stopping metric resolution." + }, + { + "left": "issue:43054", + "right": "issue:43493", + "accept": false, + "reason": "Both concern SigLIP2, but one is embedding quality and the other is implementation discrepancy against JAX." + }, + { + "left": "issue:43975", + "right": "issue:45520", + "accept": false, + "reason": "Tokenizer detokenization bug and Python 3.13 flash_attn import failure are unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44488", + "accept": false, + "reason": "Load-time noise and a specific tokenizer load failure for cjvt/sleng-bert are not the same bug." + }, + { + "left": "issue:42947", + "right": "issue:43856", + "accept": false, + "reason": "LoRA checkpointing behavior and MoE training memory usage are separate performance issues." + }, + { + "left": "issue:43643", + "right": "issue:45020", + "accept": false, + "reason": "Missing fields in AutoConfig with trust_remote_code is related to remote-code handling but not the same concrete failure as broad model-loading breakage." + }, + { + "left": "issue:41762", + "right": "issue:42915", + "accept": false, + "reason": "Gemma3 ZeRO-3 loading crash and Qwen3Moe FineGrainedFP8Config failure are different model/config combinations." + }, + { + "left": "issue:45685", + "right": "issue:45823", + "accept": false, + "reason": "An MPS histogram op error in MoE is unrelated to Gemma4 device placement mismatch." + }, + { + "left": "issue:43881", + "right": "issue:43883", + "accept": false, + "reason": "glm-4v-9b loading failure and Molmo missing all_tied_weights_keys are distinct loader bugs." + }, + { + "left": "issue:44991", + "right": "issue:45399", + "accept": false, + "reason": "Tokenizer loading for a specific model and flash-attn fallback selection are unrelated." + }, + { + "left": "issue:43824", + "right": "issue:43872", + "accept": false, + "reason": "Missing Qwen2.5-VL import and bitsandbytes Int8Params incompatibility are different integration failures." + }, + { + "left": "issue:45020", + "right": "issue:45823", + "accept": false, + "reason": "Remote_code loading regressions are not the same as Gemma4 forward device-map mismatch." + }, + { + "left": "issue:44661", + "right": "issue:45020", + "accept": false, + "reason": "A tokenizer-mapping edge case in add-new-model-like is narrower and different from the broader remote_code loading regression." + }, + { + "left": "issue:43828", + "right": "issue:45823", + "accept": false, + "reason": "Dtype mismatch under autocast and Gemma4 device mismatch are separate runtime placement issues." + }, + { + "left": "issue:42491", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility across versions is not the same as a warning about tie_word_embeddings in a different Qwen model." + }, + { + "left": "issue:45081", + "right": "issue:45399", + "accept": false, + "reason": "Mistral regex patch crash and flash-attn fallback gating are unrelated tokenizer vs attention-selection bugs." + }, + { + "left": "issue:45715", + "right": "issue:46315", + "accept": false, + "reason": "Both are tokenizer performance issues, but one is convert_ids_to_tokens and the other is convert_tokens_to_ids; not the same path." + }, + { + "left": "issue:45488", + "right": "issue:46489", + "accept": false, + "reason": "Both involve DeepSeek tokenizers, but this is a ByteLevel/Metaspace override issue while the other is an output regression from a different fix gap." + }, + { + "left": "issue:43782", + "right": "issue:44960", + "accept": false, + "reason": "Qwen3VL weight_only loading error and GLM5 are unrelated model-load problems." + }, + { + "left": "issue:45020", + "right": "issue:45356", + "accept": false, + "reason": "Remote_code breakage is broader than the specific Kimi-K2.5 tokenizer codec and warning regression." + }, + { + "left": "issue:43742", + "right": "issue:45399", + "accept": false, + "reason": "MobileLLM key error and flash-attn fallback selection do not share the same underlying bug." + }, + { + "left": "issue:43404", + "right": "issue:43950", + "accept": false, + "reason": "A tied lm_head issue in Mistral3 is not the same as corrupting non-persistent buffers during from_pretrained." + } + ] + }, + "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": 20, + "review_unit_index": 5, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:24643", + "issue:30064", + "issue:36331", + "issue:41093", + "issue:41553", + "issue:42915", + "issue:42947", + "issue:43257", + "issue:43366", + "issue:43531", + "issue:43550", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43792", + "issue:43827", + "issue:43828", + "issue:43854", + "issue:43872", + "issue:43940", + "issue:43950", + "issue:44220", + "issue:44291", + "issue:44479", + "issue:44488", + "issue:44509", + "issue:44530", + "issue:44560", + "issue:44589", + "issue:44805", + "issue:44863", + "issue:44910", + "issue:44918", + "issue:44928", + "issue:44991", + "issue:45305", + "issue:45313", + "issue:45325", + "issue:45588", + "issue:45663", + "issue:45685", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:45902", + "issue:45923", + "issue:46402", + "issue:46489" + ], + "soft_pairs": [ + "issue:43827|issue:44509", + "issue:43792|issue:44220", + "issue:42915|issue:43872", + "issue:43872|issue:44291", + "issue:42947|issue:45923", + "issue:41553|issue:44991", + "issue:36331|issue:41093", + "issue:43940|issue:45313", + "issue:43257|issue:45305", + "issue:43854|issue:44863", + "issue:44928|issue:45588", + "issue:44479|issue:44560", + "issue:41093|issue:44805", + "issue:41553|issue:44488", + "issue:44863|issue:46402", + "issue:24643|issue:30064", + "issue:44530|issue:44863", + "issue:43366|issue:43828", + "issue:45325|issue:45902", + "issue:45663|issue:45823", + "issue:43366|issue:43531", + "issue:43716|issue:44560", + "issue:43950|issue:45834", + "issue:43366|issue:44589", + "issue:44560|issue:45685", + "issue:43645|issue:43646", + "issue:43550|issue:44910", + "issue:45812|issue:46489", + "issue:44560|issue:44918" + ], + "prepared_review_unit_hash": "7875b63d595d524e8f7606193418dba56e624bccb4357811c9c330c7eeb3d97c", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36649, + "estimated_input_tokens": 9163, + "estimated_eval_tokens": 18582 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct model- or feature-specific regressions. The only clear duplicate families here are the custom-model initialization regression, the removed docs pipeline-task references, and the repeated mask/tensor shape-mismatch reports.", + "confidence": 0.64, + "canonical_issue_reason": "Issue 43646 is the best umbrella target: it states the broad Transformers 5.0 custom model initialization regression, and 43645 reads like the narrower Jupyter-notebook variant of the same bug.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43646 is the cleanest representative issue to keep because it is broad, specific enough to identify the regression, and likely subsumes the more specific notebook report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43827", + "right": "issue:44509", + "accept": true, + "reason": "Same docs regression: both say removed pipeline task names are still referenced after v5." + }, + { + "left": "issue:43792", + "right": "issue:44220", + "accept": false, + "reason": "Related audio/Whisper area, but the titles do not show the same concrete failing code path." + }, + { + "left": "issue:42915", + "right": "issue:43872", + "accept": false, + "reason": "Different problems: FineGrainedFP8Config vs bitsandbytes Int8Params initialization." + }, + { + "left": "issue:43872", + "right": "issue:44291", + "accept": false, + "reason": "Same symptom keyword, but different code paths and different load mechanisms; too weak to merge." + }, + { + "left": "issue:42947", + "right": "issue:45923", + "accept": false, + "reason": "One is gradient-checkpointing behavior, the other is a missing capability flag; not the same bug." + }, + { + "left": "issue:41553", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-load failures, but for different models and different reported failures." + }, + { + "left": "issue:36331", + "right": "issue:41093", + "accept": false, + "reason": "Unrelated errors: custom trainer signature mismatch vs tensor mask shape mismatch." + }, + { + "left": "issue:43940", + "right": "issue:45313", + "accept": false, + "reason": "Both are ZeRO-3 weight-loading issues, but they target different Qwen model families and likely different fixes." + }, + { + "left": "issue:43257", + "right": "issue:45305", + "accept": false, + "reason": "Different DeepSpeed/accelerate problems: missing MoE weight conversion vs gradient averaging with loss kwargs." + }, + { + "left": "issue:43854", + "right": "issue:44863", + "accept": false, + "reason": "Different models and different loading failures; no clear shared underlying defect." + }, + { + "left": "issue:44928", + "right": "issue:45588", + "accept": false, + "reason": "Both touch attention/positioning, but one is RLHF gradient instability and the other is a flash-attention AttributeError." + }, + { + "left": "issue:44479", + "right": "issue:44560", + "accept": false, + "reason": "Both are video-related Qwen regressions, but the reported failure modes are different." + }, + { + "left": "issue:41093", + "right": "issue:44805", + "accept": true, + "reason": "Same mask/tensor shape-mismatch error signature; 44805 looks like a later duplicate report of the same bug class." + }, + { + "left": "issue:41553", + "right": "issue:44488", + "accept": false, + "reason": "Different tokenizer/model-loading issues with no evidence they share the same code path." + }, + { + "left": "issue:44863", + "right": "issue:46402", + "accept": false, + "reason": "Different models and different loading symptoms; not the same underlying issue." + }, + { + "left": "issue:24643", + "right": "issue:30064", + "accept": false, + "reason": "Training with DeepSpeed weight-dimension error is unrelated to void segmentation-map preprocessing." + }, + { + "left": "issue:44530", + "right": "issue:44863", + "accept": false, + "reason": "PagedAttentionCache crash vs checkpoint-loading failure; unrelated." + }, + { + "left": "issue:43366", + "right": "issue:43828", + "accept": false, + "reason": "GGUF gpt-oss support and autocast dtype mismatch are unrelated feature/bug reports." + }, + { + "left": "issue:45325", + "right": "issue:45902", + "accept": false, + "reason": "Different Qwen3.x rotary-position bugs with different symptoms and likely different fixes." + }, + { + "left": "issue:45663", + "right": "issue:45823", + "accept": false, + "reason": "Both are Gemma4, but one is a per-layer shared_kv_states KeyError and the other is a device-map mismatch." + }, + { + "left": "issue:43366", + "right": "issue:43531", + "accept": false, + "reason": "GGUF gpt-oss support and Qwen3-MoE sliding_window behavior are not the same issue." + }, + { + "left": "issue:43716", + "right": "issue:44560", + "accept": false, + "reason": "Different multimodal regressions: image-preprocessor/model dtype mismatch vs video StopIteration." + }, + { + "left": "issue:43950", + "right": "issue:45834", + "accept": false, + "reason": "Non-persistent buffer corruption is unrelated to the Kosmos2.5 OCR index error." + }, + { + "left": "issue:43366", + "right": "issue:44589", + "accept": false, + "reason": "Unrelated model-support and storage-type errors." + }, + { + "left": "issue:43645", + "right": "issue:43646", + "accept": true, + "reason": "Same underlying custom-model initialization regression; 43645 is just the Jupyter-specific wording of the broader 43646 report." + }, + { + "left": "issue:43550", + "right": "issue:44910", + "accept": false, + "reason": "Different attention/kernel failures on different models and different symptoms." + }, + { + "left": "issue:45812", + "right": "issue:46489", + "accept": false, + "reason": "Different tokenizer regressions for different model families and different change histories." + }, + { + "left": "issue:44560", + "right": "issue:45685", + "accept": false, + "reason": "Video model StopIteration and MPS histogram int support are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The accepted soft pairs are supported by the titles and relative specificity (43827/44509, 41093/44805, 43645/43646), and the rejected pairs avoid over-merging on weak symptom similarity. No obvious overstatement." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 6, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36010", + "issue:38175", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42617", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43366", + "issue:43596", + "issue:43632", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43708", + "issue:43720", + "issue:43824", + "issue:43854", + "issue:43856", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44075", + "issue:44155", + "issue:44387", + "issue:44451", + "issue:44488", + "issue:44534", + "issue:44560", + "issue:44661", + "issue:44756", + "issue:44805", + "issue:44877", + "issue:44960", + "issue:44977", + "issue:45030", + "issue:45276", + "issue:45313", + "issue:45357", + "issue:45488", + "issue:45507", + "issue:45584", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:46468" + ], + "soft_pairs": [ + "issue:42915|issue:43957", + "issue:43824|issue:43957", + "issue:42617|issue:43366", + "issue:36010|issue:42915", + "issue:44661|issue:45823", + "issue:44560|issue:44805", + "issue:38175|issue:43994", + "issue:43632|issue:45834", + "issue:44155|issue:45584", + "issue:43638|issue:44661", + "issue:43596|issue:44661", + "issue:44877|issue:45030", + "issue:44661|issue:44805", + "issue:43317|issue:44756", + "issue:38175|issue:39692", + "issue:42491|issue:43720", + "issue:41720|issue:42915", + "issue:44451|issue:44488", + "issue:45313|issue:45357", + "issue:44960|issue:44977", + "issue:42491|issue:45276", + "issue:42915|issue:44756", + "issue:43644|issue:44534", + "issue:45507|issue:46468", + "issue:39692|issue:43994", + "issue:43708|issue:44075", + "issue:43856|issue:43873", + "issue:43644|issue:43645", + "issue:43854|issue:43931", + "issue:43856|issue:44387", + "issue:43257|issue:43854", + "issue:45488|issue:45812" + ], + "prepared_review_unit_hash": "b661dd0f6079909b3a20e65bc27c4cf0f74f999b04344624b2d4639a6fb33ba9", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37089, + "estimated_input_tokens": 9273, + "estimated_eval_tokens": 18802 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab-bag of unrelated Transformers v5/model-specific regressions. Only one soft edge is a clear duplicate; the rest are too cross-model or only share a broad subsystem like loading, quantization, or device mapping.", + "confidence": 0.81, + "canonical_issue_reason": "issue:43644 is the clearest canonical issue: it is earlier, broadly phrased, and has the exact duplicate/rewording in issue:44534.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43644 is the best representative of the duplicate set because it captures the shared concrete bug and is the earliest instance.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:43957", + "accept": false, + "reason": "Both are loading failures, but one is Qwen3Moe FP8-specific and the other is a generic meta-device loading regression; not the same concrete bug." + }, + { + "left": "issue:43824", + "right": "issue:43957", + "accept": false, + "reason": "Import error for a specific class vs a model-loading/meta-device failure; unrelated code paths." + }, + { + "left": "issue:42617", + "right": "issue:43366", + "accept": false, + "reason": "3d_parallel.py runtime issue and GGUF/gpt-oss support are different problems." + }, + { + "left": "issue:36010", + "right": "issue:42915", + "accept": false, + "reason": "GenerationMixin import failure is unrelated to Qwen3Moe FP8 loading." + }, + { + "left": "issue:44661", + "right": "issue:45823", + "accept": false, + "reason": "Tokenizer-mapping bug and Gemma4 device mismatch are different failures." + }, + { + "left": "issue:44560", + "right": "issue:44805", + "accept": false, + "reason": "Different runtime errors in different model paths; only a broad 'indexing error' resemblance." + }, + { + "left": "issue:38175", + "right": "issue:43994", + "accept": false, + "reason": "Both involve SigLIP2, but the reported symptoms and likely failure modes differ enough to avoid merging without stronger evidence." + }, + { + "left": "issue:43632", + "right": "issue:45834", + "accept": false, + "reason": "Transformer v5 flag regression and Kosmos2.5 OCR index error are unrelated." + }, + { + "left": "issue:44155", + "right": "issue:45584", + "accept": false, + "reason": "AudioFlamingo3 cross-track leakage and Whisper empty-transcription failure are different bugs." + }, + { + "left": "issue:43638", + "right": "issue:44661", + "accept": false, + "reason": "Bert/DeepSpeed zero3 initialization error is unrelated to the tokenizer-mapping add-new-model-like issue." + }, + { + "left": "issue:43596", + "right": "issue:44661", + "accept": false, + "reason": "Both are load-time failures, but they affect different mechanisms and model families." + }, + { + "left": "issue:44877", + "right": "issue:45030", + "accept": false, + "reason": "Config-validation failures for different models; not the same concrete regression." + }, + { + "left": "issue:44661", + "right": "issue:44805", + "accept": false, + "reason": "Tokenizer-mapping bug and mask/tensor shape mismatch are not the same issue." + }, + { + "left": "issue:43317", + "right": "issue:44756", + "accept": false, + "reason": "Dequantized offload failure vs mmap/OOM on Strix Halo; too different to treat as duplicates." + }, + { + "left": "issue:38175", + "right": "issue:39692", + "accept": false, + "reason": "Same model family, but one is a zero-probability symptom and the other is a docs example with multiple issues, including quantization; not clearly one bug." + }, + { + "left": "issue:42491", + "right": "issue:43720", + "accept": false, + "reason": "Qwen3 MoE LoRA compatibility and BitNet packed-weight loading are distinct code paths." + }, + { + "left": "issue:41720", + "right": "issue:42915", + "accept": false, + "reason": "Both are Qwen-related failures, but one is CUDA assert under device mapping and the other is FP8 config loading; not the same bug." + }, + { + "left": "issue:44451", + "right": "issue:44488", + "accept": false, + "reason": "Similar 'cannot load model' wording, but different models and insufficient evidence of a single shared root cause." + }, + { + "left": "issue:45313", + "right": "issue:45357", + "accept": false, + "reason": "ZeRO-3 weight loading and save_pretrained key regression are opposite sides of the workflow and not the same bug." + }, + { + "left": "issue:44960", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and different failure modes." + }, + { + "left": "issue:42491", + "right": "issue:45276", + "accept": false, + "reason": "LoRA/load-compatibility issue vs embedding resize propagation bug; unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3Moe FP8 failure and Strix Halo mmap OOM share only a vague memory-loading theme." + }, + { + "left": "issue:43644", + "right": "issue:44534", + "accept": true, + "reason": "Same bug: identical title/body theme about non-persistent buffers being filled with junk in Transformers v5." + }, + { + "left": "issue:45507", + "right": "issue:46468", + "accept": false, + "reason": "GraniteMoEHybrid invalid-method calls and Nemotron example breakage are different issues." + }, + { + "left": "issue:39692", + "right": "issue:43994", + "accept": false, + "reason": "The documentation example bug and the nonsensical-output report are related at best by model family, but not clearly the same concrete defect." + }, + { + "left": "issue:43708", + "right": "issue:44075", + "accept": false, + "reason": "Trainer checkpoint math and SGD argument handling are different trainer bugs." + }, + { + "left": "issue:43856", + "right": "issue:43873", + "accept": false, + "reason": "Both mention quantization/offloading/memory, but one is training memory inefficiency and the other is offload not working as expected." + }, + { + "left": "issue:43644", + "right": "issue:43645", + "accept": false, + "reason": "Both are Transformers v5 regressions, but one is buffer initialization and the other is custom-model initialization in notebooks; not the same bug." + }, + { + "left": "issue:43854", + "right": "issue:43931", + "accept": false, + "reason": "Different models and different loading errors." + }, + { + "left": "issue:43856", + "right": "issue:44387", + "accept": false, + "reason": "Both concern memory, but one is Qwen3 MoE training efficiency and the other is int4 reserved-memory OOM under Transformers 5.x." + }, + { + "left": "issue:43257", + "right": "issue:43854", + "accept": false, + "reason": "Qwen3 MoE weight conversion/loading and GLM-4.7-Flash unit-test loading are unrelated." + }, + { + "left": "issue:45488", + "right": "issue:45812", + "accept": false, + "reason": "Different tokenizer regressions affecting different model families; same broad area, not the same concrete bug." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The lone positive edge (issue:43644 \u2194 issue:44534) is well grounded by near-identical wording and the earlier issue is a reasonable canonical representative. The remaining soft edges are phrased conservatively, generally distinguishing model-specific or code-path-specific failures without overclaiming duplication. Overall, the summary stays within the evidence in the packet." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 7, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36010", + "issue:39404", + "issue:39692", + "issue:41720", + "issue:42491", + "issue:42947", + "issue:43335", + "issue:43381", + "issue:43404", + "issue:43425", + "issue:43454", + "issue:43854", + "issue:43873", + "issue:43931", + "issue:43957", + "issue:44155", + "issue:44164", + "issue:44291", + "issue:44292", + "issue:44387", + "issue:44402", + "issue:44479", + "issue:44488", + "issue:44756", + "issue:44863", + "issue:44877", + "issue:44910", + "issue:44912", + "issue:44928", + "issue:44960", + "issue:44991", + "issue:45216", + "issue:45276", + "issue:45292", + "issue:45335", + "issue:45362", + "issue:45406", + "issue:45588", + "issue:45663", + "issue:45684", + "issue:45685", + "issue:45698", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45859", + "issue:46050", + "issue:46468" + ], + "soft_pairs": [ + "issue:42491|issue:45292", + "issue:42491|issue:43931", + "issue:44488|issue:44991", + "issue:45276|issue:45335", + "issue:44910|issue:45588", + "issue:44164|issue:44291", + "issue:45362|issue:45406", + "issue:43404|issue:43454", + "issue:43425|issue:43957", + "issue:39692|issue:44960", + "issue:44402|issue:44488", + "issue:42947|issue:45704", + "issue:36010|issue:39404", + "issue:45663|issue:45759", + "issue:43873|issue:44387", + "issue:45663|issue:45859", + "issue:44155|issue:45823", + "issue:44291|issue:44756", + "issue:44292|issue:44912", + "issue:44960|issue:46050", + "issue:44387|issue:44756", + "issue:43335|issue:45663", + "issue:42947|issue:43381", + "issue:45276|issue:45292", + "issue:44863|issue:46468", + "issue:45685|issue:46468", + "issue:43854|issue:45216", + "issue:45684|issue:45698", + "issue:44479|issue:44877", + "issue:43381|issue:44928", + "issue:44164|issue:44756", + "issue:43957|issue:44756", + "issue:45588|issue:45663", + "issue:41720|issue:44155" + ], + "prepared_review_unit_hash": "93f370129fe54111b069fe41aaff631084478123fc53f675f4527f1995327105", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 37964, + "estimated_input_tokens": 9491, + "estimated_eval_tokens": 19238 + }, + "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 bugs. A few are about save/load or token embedding propagation, but the soft pairs do not look like true duplicates; they mostly share a subsystem, not the same underlying defect.", + "confidence": 0.82, + "canonical_issue_reason": "Issue 45698 is the best single representative because it is the broadest active save/load regression in the set and is closest in theme to several neighboring custom-module and save_pretrained/from_pretrained reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45698 is the most representative cluster anchor: it is general, recent, open, and centered on save/load behavior rather than a narrow model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42491", + "right": "issue:45292", + "accept": false, + "reason": "Both involve model/config loading and embedding-related behavior, but one is a Qwen3 MoE LoRA compatibility problem and the other is a resize_token_embeddings propagation bug." + }, + { + "left": "issue:42491", + "right": "issue:43931", + "accept": false, + "reason": "Different models and failures: Qwen3 MoE LoRA incompatibility vs Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:44488", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer/model load failures, but they affect different models and describe different concrete breakages." + }, + { + "left": "issue:45276", + "right": "issue:45335", + "accept": false, + "reason": "Same general API area, but Gemma4-specific missing propagation to multiple embeddings is not clearly the same bug as the T5Gemma decoder embed_tokens issue." + }, + { + "left": "issue:44910", + "right": "issue:45588", + "accept": false, + "reason": "Both are flash-attention related crashes, but the Qwen3.5 position_ids bug and the sink-less s_aux=None AttributeError are different code paths." + }, + { + "left": "issue:44164", + "right": "issue:44291", + "accept": false, + "reason": "Both touch model loading/saving, but extra_state handling and init_empty_weights/_is_hf_initialized are separate bugs." + }, + { + "left": "issue:45362", + "right": "issue:45406", + "accept": false, + "reason": "Different surfaces and failures: Qwen3.5 chat crash vs Gemma4Processor missing _tokenizer in transformers serve." + }, + { + "left": "issue:43404", + "right": "issue:43454", + "accept": false, + "reason": "Both mention untied lm_head, but they are different model implementations and likely separate missing-tie bugs." + }, + { + "left": "issue:43425", + "right": "issue:43957", + "accept": false, + "reason": "Torch 2.10 compatibility and meta-device loading failures are not the same underlying issue." + }, + { + "left": "issue:39692", + "right": "issue:44960", + "accept": false, + "reason": "A broken documentation example and a GLM5 loading issue are unrelated." + }, + { + "left": "issue:44402", + "right": "issue:44488", + "accept": false, + "reason": "Both are tokenizer-related, but one is vocab size mismatch and the other is a model load failure for a different checkpoint." + }, + { + "left": "issue:42947", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing/LoRA and an apex FusedRMSNorm memory leak are unrelated bugs." + }, + { + "left": "issue:36010", + "right": "issue:39404", + "accept": false, + "reason": "ImportError for GenerationMixin and Whisper pipeline return_language regression are unrelated." + }, + { + "left": "issue:45663", + "right": "issue:45759", + "accept": false, + "reason": "Gemma4 FSDP2 shared_kv_states KeyError and Qwen composite-config unwrapping are different model-specific issues." + }, + { + "left": "issue:43873", + "right": "issue:44387", + "accept": false, + "reason": "Both involve memory behavior under quantization, but one is generic offloading and the other is int4 reserved-memory/OOM growth." + }, + { + "left": "issue:45663", + "right": "issue:45859", + "accept": false, + "reason": "Gemma4 training FSDP2 bug and Qwen3_5Moe CPU-offload incompatibility are different failures." + }, + { + "left": "issue:44155", + "right": "issue:45823", + "accept": false, + "reason": "AudioFlamingo3 batched inference leakage is unrelated to Gemma4 device-map mismatch." + }, + { + "left": "issue:44291", + "right": "issue:44756", + "accept": false, + "reason": "Unexpected init_empty_weights argument and Strix Halo mmap/OOM are unrelated." + }, + { + "left": "issue:44292", + "right": "issue:44912", + "accept": false, + "reason": "init_empty_weights/_is_hf_initialized loading failure is unrelated to MXFP4 quantization fallback." + }, + { + "left": "issue:44960", + "right": "issue:46050", + "accept": false, + "reason": "GLM5 loading and sinq save/load corruption are different model-specific issues." + }, + { + "left": "issue:44387", + "right": "issue:44756", + "accept": false, + "reason": "Both mention OOM, but the causes differ: int4 reserved-memory growth vs mmap on Strix Halo." + }, + { + "left": "issue:43335", + "right": "issue:45663", + "accept": false, + "reason": "Sparse-layer creation in SwitchTransformers and Gemma4 FSDP2 shared_kv_states are unrelated." + }, + { + "left": "issue:42947", + "right": "issue:43381", + "accept": false, + "reason": "Both are gradient-checkpointing related, but one is ineffective LoRA training and the other is an eval-mode restriction." + }, + { + "left": "issue:45276", + "right": "issue:45292", + "accept": false, + "reason": "Highly similar theme, but the Gemma4-specific multi-embedding propagation bug is not clearly the same defect as the general output_embeddings issue." + }, + { + "left": "issue:44863", + "right": "issue:46468", + "accept": false, + "reason": "NemotronH checkpoint loading and the official Nemotron example failure are related by model family only, not the same bug." + }, + { + "left": "issue:45685", + "right": "issue:46468", + "accept": false, + "reason": "An MPS histogram kernel limitation is unrelated to the Nemotron example loading breakage." + }, + { + "left": "issue:43854", + "right": "issue:45216", + "accept": false, + "reason": "GLM-4.7-Flash unit-test loading and Qwen3.5 save_pretrained regression are separate model-specific bugs." + }, + { + "left": "issue:45684", + "right": "issue:45698", + "accept": false, + "reason": "Both are save_pretrained-related, but one is file permission propagation and the other is wrong custom-module loading." + }, + { + "left": "issue:44479", + "right": "issue:44877", + "accept": false, + "reason": "Video-input regression and strict Granite Speech config loading are different issues." + }, + { + "left": "issue:43381", + "right": "issue:44928", + "accept": false, + "reason": "Eval-mode gradient checkpointing restriction and Qwen3.5 RLHF NaN explosion are not the same bug." + }, + { + "left": "issue:44164", + "right": "issue:44756", + "accept": false, + "reason": "save/from_pretrained extra_state handling is unrelated to mmap-induced OOM on Strix Halo." + }, + { + "left": "issue:43957", + "right": "issue:44756", + "accept": false, + "reason": "Meta-device loading failures and Strix Halo mmap OOM are different code paths." + }, + { + "left": "issue:45588", + "right": "issue:45663", + "accept": false, + "reason": "Flash-attention sink-less-model crash and Gemma4 FSDP2 KeyError are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:44155", + "accept": false, + "reason": "Qwen3 auto device-map cuda assert and AudioFlamingo3 batched track leakage 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": 20, + "review_unit_index": 8, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36032", + "issue:39404", + "issue:42796", + "issue:43299", + "issue:43404", + "issue:43425", + "issue:43452", + "issue:43596", + "issue:43611", + "issue:43645", + "issue:43646", + "issue:43708", + "issue:43716", + "issue:43828", + "issue:43856", + "issue:43901", + "issue:43906", + "issue:43931", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44164", + "issue:44265", + "issue:44402", + "issue:44451", + "issue:44479", + "issue:44509", + "issue:44530", + "issue:44589", + "issue:44671", + "issue:44683", + "issue:44756", + "issue:44805", + "issue:44898", + "issue:44912", + "issue:44918", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45292", + "issue:45335", + "issue:45701", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46315", + "issue:46326", + "issue:46491" + ], + "soft_pairs": [ + "issue:43646|issue:43708", + "issue:45292|issue:45335", + "issue:43645|issue:44164", + "issue:43828|issue:43957", + "issue:46315|issue:46491", + "issue:44402|issue:44451", + "issue:43425|issue:43828", + "issue:45812|issue:46315", + "issue:44589|issue:44756", + "issue:44683|issue:45823", + "issue:43611|issue:43645", + "issue:43425|issue:45823", + "issue:43425|issue:43716", + "issue:44530|issue:45084", + "issue:39404|issue:45701", + "issue:42796|issue:46326", + "issue:43906|issue:44530", + "issue:44265|issue:44898", + "issue:44589|issue:45084", + "issue:44479|issue:44805", + "issue:43901|issue:44509", + "issue:43425|issue:44918", + "issue:39404|issue:44912", + "issue:43596|issue:44683", + "issue:36032|issue:43452", + "issue:44164|issue:45759", + "issue:43299|issue:43931", + "issue:44991|issue:45081", + "issue:44671|issue:44991", + "issue:43994|issue:44079", + "issue:43856|issue:44756", + "issue:42796|issue:43404" + ], + "prepared_review_unit_hash": "789374209ffa92eb8154344446f1914253b1c7feeca1d6ae435f4307e4ce88f4", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37476, + "estimated_input_tokens": 9369, + "estimated_eval_tokens": 18994 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous regression reports around tokenizers/model loading, with one clear duplicate pair about `resize_token_embeddings` not propagating to all tied output embeddings. I did not find any PRs here.", + "confidence": 0.61, + "canonical_issue_reason": "issue:45292 is the broadest statement of the `resize_token_embeddings` propagation bug; issue:45335 is a model-specific instance of the same missing update to downstream embeddings.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45292 best represents the underlying bug because it describes the general embedding-resize failure, rather than a single model variant.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43646", + "right": "issue:43708", + "accept": false, + "reason": "Different failures: custom model init/loading vs Trainer checkpoint resume step counting." + }, + { + "left": "issue:45292", + "right": "issue:45335", + "accept": true, + "reason": "Same underlying bug: `resize_token_embeddings` updates the main embeddings but not the tied/output embeddings; 45335 is a t5gemma-specific instance." + }, + { + "left": "issue:43645", + "right": "issue:44164", + "accept": false, + "reason": "Notebook custom-model init regression is unrelated to `extra_state` save/load handling." + }, + { + "left": "issue:43828", + "right": "issue:43957", + "accept": false, + "reason": "Both are dtype-related, but one is autocast inference and the other is meta-device loading; different code paths." + }, + { + "left": "issue:46315", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions: slow `convert_tokens_to_ids` performance vs CodeLlama round-trip spacing." + }, + { + "left": "issue:44402", + "right": "issue:44451", + "accept": false, + "reason": "A specific tokenizer vocab/special-token issue vs a generic model loading failure; not clearly the same bug." + }, + { + "left": "issue:43425", + "right": "issue:43828", + "accept": false, + "reason": "Torch compatibility report vs autocast dtype mismatch; unrelated problems." + }, + { + "left": "issue:45812", + "right": "issue:46315", + "accept": false, + "reason": "Granite token ID regression and slow `convert_tokens_to_ids` performance are different tokenizer bugs." + }, + { + "left": "issue:44589", + "right": "issue:44756", + "accept": false, + "reason": "Float8 storage lookup failure vs mmap/OOM on Strix Halo; different symptoms and causes." + }, + { + "left": "issue:44683", + "right": "issue:45823", + "accept": false, + "reason": "Compiled flex_attention on torch>=2.9 is unrelated to Gemma4 device-mismatch during forward." + }, + { + "left": "issue:43611", + "right": "issue:43645", + "accept": false, + "reason": "Both involve model loading, but one is `base_model_prefix` handling and the other is custom model initialization in notebooks." + }, + { + "left": "issue:43425", + "right": "issue:45823", + "accept": false, + "reason": "Torch 2.10 incompatibility is not the same as a Gemma4 device-map forward mismatch." + }, + { + "left": "issue:43425", + "right": "issue:43716", + "accept": false, + "reason": "General torch incompatibility vs a specific image-preprocessor/model dtype mismatch." + }, + { + "left": "issue:44530", + "right": "issue:45084", + "accept": false, + "reason": "PagedAttentionCache linear_attention crash and non-template-node compilation error are different issues." + }, + { + "left": "issue:39404", + "right": "issue:45701", + "accept": false, + "reason": "Whisper `return_language` pipeline behavior is not the same as a broad tokenization change report." + }, + { + "left": "issue:42796", + "right": "issue:46326", + "accept": false, + "reason": "Ministral BPE-marker handling and `continue_final_message` for mistral_common are related to Mistral tokenization but not the same concrete bug." + }, + { + "left": "issue:43906", + "right": "issue:44530", + "accept": false, + "reason": "Isolated reproduction of a prior issue vs a Qwen3.5 PagedAttentionCache crash; no direct match." + }, + { + "left": "issue:44265", + "right": "issue:44898", + "accept": false, + "reason": "torch.export compatibility and Perceiver interpolation failure are unrelated model/export bugs." + }, + { + "left": "issue:44589", + "right": "issue:45084", + "accept": false, + "reason": "Different failure modes: missing Float8 storage object vs template-node compilation error." + }, + { + "left": "issue:44479", + "right": "issue:44805", + "accept": false, + "reason": "Video-input regression for Qwen VL models vs a mask shape IndexError; not the same bug." + }, + { + "left": "issue:43901", + "right": "issue:44509", + "accept": false, + "reason": "Both are docs regressions, but they concern different removed/changed pipeline APIs." + }, + { + "left": "issue:43425", + "right": "issue:44918", + "accept": false, + "reason": "Torch incompatibility is unrelated to TRL SFT unpacking of Qwen3.5 embeddings." + }, + { + "left": "issue:39404", + "right": "issue:44912", + "accept": false, + "reason": "Whisper pipeline language return and MXFP4 quantization fallback are different regressions." + }, + { + "left": "issue:43596", + "right": "issue:44683", + "accept": false, + "reason": "Deepspeed zero3/BertModel index error is unrelated to flex_attention compilation on newer torch." + }, + { + "left": "issue:36032", + "right": "issue:43452", + "accept": false, + "reason": "T5 tokenizer load conflict and `gguf_file` loading breakage are different tokenizer/model-loading bugs." + }, + { + "left": "issue:44164", + "right": "issue:45759", + "accept": false, + "reason": "`extra_state` serialization and composite Qwen config unwrapping are separate loading paths." + }, + { + "left": "issue:43299", + "right": "issue:43931", + "accept": false, + "reason": "Same model family, but one is a broad v5 dev loading regression and the other is a specific weight-shape mismatch; not clearly the same concrete bug." + }, + { + "left": "issue:44991", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer load failure for est-roberta and Mistral regex patch crash are unrelated." + }, + { + "left": "issue:44671", + "right": "issue:44991", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and tokenizer loading failure for a different model are distinct issues." + }, + { + "left": "issue:43994", + "right": "issue:44079", + "accept": false, + "reason": "Siglip2 wrong outputs and `ModelOutput` key assignment are not the same code-path problem." + }, + { + "left": "issue:43856", + "right": "issue:44756", + "accept": false, + "reason": "Qwen3 MoE training memory usage and disabling mmap for Strix Halo OOM are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:43404", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to Mistral3 lm_head weight tying." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet, and the only accepted soft pair (issue:45292 / issue:45335) is a conservative duplicate-style match on the same resize-token-embeddings propagation bug. The rest of the soft-edge verdicts stay appropriately cautious, and it\u2019s correct that no PRs were present." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 9, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:32090", + "issue:34567", + "issue:36032", + "issue:39404", + "issue:40444", + "issue:40990", + "issue:42175", + "issue:42915", + "issue:43317", + "issue:43421", + "issue:43452", + "issue:43482", + "issue:43541", + "issue:43582", + "issue:43611", + "issue:43632", + "issue:43643", + "issue:43644", + "issue:43645", + "issue:43646", + "issue:43716", + "issue:43883", + "issue:43906", + "issue:44164", + "issue:44206", + "issue:44220", + "issue:44451", + "issue:44462", + "issue:44479", + "issue:44488", + "issue:44530", + "issue:44661", + "issue:44671", + "issue:44683", + "issue:44737", + "issue:44811", + "issue:44987", + "issue:45081", + "issue:45084", + "issue:45507", + "issue:45584", + "issue:45859", + "issue:46050", + "issue:46178", + "issue:46285", + "issue:46326", + "issue:46402", + "issue:46489" + ], + "soft_pairs": [ + "issue:44462|issue:46489", + "issue:44164|issue:44479", + "issue:43906|issue:46326", + "issue:44737|issue:45859", + "issue:44661|issue:44683", + "issue:44488|issue:45081", + "issue:39404|issue:44671", + "issue:43611|issue:43644", + "issue:46285|issue:46402", + "issue:43716|issue:45084", + "issue:43582|issue:43632", + "issue:40990|issue:43421", + "issue:44488|issue:44987", + "issue:32090|issue:39404", + "issue:36032|issue:42915", + "issue:43643|issue:44530", + "issue:46050|issue:46326", + "issue:34567|issue:40444", + "issue:44451|issue:44987", + "issue:43541|issue:45084", + "issue:44811|issue:45584", + "issue:42915|issue:44530", + "issue:43644|issue:43646", + "issue:43452|issue:43482", + "issue:43883|issue:44488", + "issue:44206|issue:44220", + "issue:43317|issue:43482", + "issue:36032|issue:43645", + "issue:45084|issue:45507", + "issue:42175|issue:46178" + ], + "prepared_review_unit_hash": "c1bd690c3337cce6ad5d270afc45f1c4d7da197ec3bef47e9a51abcd5fe2d544", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36648, + "estimated_input_tokens": 9162, + "estimated_eval_tokens": 18580 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a grab bag of unrelated issues; only one soft duplicate looks plausible. The strongest anchor is the generic GGUF-loading regression, with the model-specific Qwen2.5 report as its likely duplicate.", + "confidence": 0.76, + "canonical_issue_reason": "issue:43452 is the broadest and most generic report of the GGUF-loading failure, so it best serves as the anchor for the only plausible duplicate pair.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43452 is the clearest representative of the shared GGUF loader regression and is better suited as the canonical issue than the model-specific follow-up.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44462", + "right": "issue:46489", + "accept": false, + "reason": "Both are tokenizer regressions, but one is about repository tokenizer.json resolution while the other is a DeepSeek-Coder-specific output bug; not clearly the same code path." + }, + { + "left": "issue:44164", + "right": "issue:44479", + "accept": false, + "reason": "Unrelated failures: save/from_pretrained extra_state handling vs a video-input regression in Qwen VL models." + }, + { + "left": "issue:43906", + "right": "issue:46326", + "accept": false, + "reason": "Both mention tokenizer-related behavior, but one is an isolated repro of another issue and the other is a mistral_common continue_final_message bug; not the same defect." + }, + { + "left": "issue:44737", + "right": "issue:45859", + "accept": false, + "reason": "Both involve device/offload behavior, but they affect different models and different code paths." + }, + { + "left": "issue:44661", + "right": "issue:44683", + "accept": false, + "reason": "Different subsystems entirely: tokenizer mapping discovery vs torch 2.9 flex_attention compilation." + }, + { + "left": "issue:44488", + "right": "issue:45081", + "accept": false, + "reason": "Both are tokenizer-loading problems, but one is model-loading for a BERT repo and the other is a Mistral regex patch crash." + }, + { + "left": "issue:39404", + "right": "issue:44671", + "accept": false, + "reason": "Whisper return_language pipeline regression is unrelated to CamemBERT masked-LM prediction changes." + }, + { + "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; different bugs." + }, + { + "left": "issue:46285", + "right": "issue:46402", + "accept": false, + "reason": "Different models and symptoms: CLIP structure drift vs Sam3Model text encoder weights not loading." + }, + { + "left": "issue:43716", + "right": "issue:45084", + "accept": false, + "reason": "Unrelated: image-preprocessor dtype mismatch vs template compilation error." + }, + { + "left": "issue:43582", + "right": "issue:43632", + "accept": false, + "reason": "Different regression classes: Apple Silicon allocator warmup TypeError vs _is_hf_initialized flag breakage." + }, + { + "left": "issue:40990", + "right": "issue:43421", + "accept": false, + "reason": "Perplexity issue on GPT-OSS is unrelated to tokenizer backend post-processor updates." + }, + { + "left": "issue:44488", + "right": "issue:44987", + "accept": false, + "reason": "Both are load failures, but for different models and likely different root causes." + }, + { + "left": "issue:32090", + "right": "issue:39404", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is unrelated to Whisper pipeline language-return behavior." + }, + { + "left": "issue:36032", + "right": "issue:42915", + "accept": false, + "reason": "Tokenizer special-token method conflict vs Qwen3Moe FP8 configuration failure are unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44530", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss is unrelated to Qwen3.5 cache/group-type crashes." + }, + { + "left": "issue:46050", + "right": "issue:46326", + "accept": false, + "reason": "Quantized model save/load breakage is unrelated to the mistral_common continue_final_message issue." + }, + { + "left": "issue:34567", + "right": "issue:40444", + "accept": false, + "reason": "TrainerState token counting and Qwen2.5-VL iterable dataset failures are different bugs." + }, + { + "left": "issue:44451", + "right": "issue:44987", + "accept": false, + "reason": "Different model-loading regressions affecting unrelated repositories." + }, + { + "left": "issue:43541", + "right": "issue:45084", + "accept": false, + "reason": "Torch Dynamo grouped_mm tracing error is not the same as non-template-node compilation failure." + }, + { + "left": "issue:44811", + "right": "issue:45584", + "accept": false, + "reason": "Both are Whisper-related, but one is decode skip_special_tokens handling and the other is empty-transcription generation after align_special_tokens." + }, + { + "left": "issue:42915", + "right": "issue:44530", + "accept": false, + "reason": "Both are Qwen-family issues, but one is FP8 config setup and the other is paged attention cache/group-type handling." + }, + { + "left": "issue:43644", + "right": "issue:43646", + "accept": false, + "reason": "Different v5 regressions: junk buffers vs custom model initialization." + }, + { + "left": "issue:43452", + "right": "issue:43482", + "accept": true, + "reason": "Both point to the same GGUF loading path breaking in transformers v5; 43482 looks like a model-specific manifestation of the broader 43452 regression." + }, + { + "left": "issue:43883", + "right": "issue:44488", + "accept": false, + "reason": "Molmo tied-weights attribute error is unrelated to model-load failures for BERT repositories." + }, + { + "left": "issue:44206", + "right": "issue:44220", + "accept": false, + "reason": "Related audio feature extraction area, but one is a v5.2.0 regression from an unsupported center arg and the other is a different extract_fbank issue." + }, + { + "left": "issue:43317", + "right": "issue:43482", + "accept": false, + "reason": "Dequantized-model device_map offload failure is not the same as a GGUF loading regression." + }, + { + "left": "issue:36032", + "right": "issue:43645", + "accept": false, + "reason": "Tokenizer add_special_tokens conflict and Jupyter custom model initialization are unrelated." + }, + { + "left": "issue:45084", + "right": "issue:45507", + "accept": false, + "reason": "Different problems: template compilation error vs GraniteMoEHybrid calling an invalid method." + }, + { + "left": "issue:42175", + "right": "issue:46178", + "accept": false, + "reason": "TensorFlow backend packaging issue is unrelated to tensorflow_text backend mapping validation." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted soft edge\u2014issue:43452 with issue:43482\u2014is reasonably supported by the titles and the broad GGUF-loading regression framing. The rest of the negative verdicts are cautious and do not overclaim overlap." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 10, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "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:43611", + "issue:43645", + "issue:43646", + "issue:43742", + "issue:43819", + "issue:43883", + "issue:43994", + "issue:44062", + "issue:44220", + "issue:44265", + "issue:44291", + "issue:44403", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44534", + "issue:44589", + "issue:44610", + "issue:44683", + "issue:44849", + "issue:44877", + "issue:44987", + "issue:44991", + "issue:45081", + "issue:45084", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45812", + "issue:45987", + "issue:46315", + "issue:46326", + "issue:46491" + ], + "soft_pairs": [ + "issue:44849|issue:45325", + "issue:44464|issue:45987", + "issue:43262|issue:43377", + "issue:44987|issue:44991", + "issue:43646|issue:44534", + "issue:43883|issue:44991", + "issue:43404|issue:44466", + "issue:44403|issue:45081", + "issue:42915|issue:45084", + "issue:43994|issue:44610", + "issue:43377|issue:43819", + "issue:43322|issue:43540", + "issue:28530|issue:34567", + "issue:43645|issue:44534", + "issue:43089|issue:45987", + "issue:44451|issue:45081", + "issue:44987|issue:45081", + "issue:44265|issue:44610", + "issue:44220|issue:44683", + "issue:44991|issue:46315", + "issue:45216|issue:45357", + "issue:42915|issue:44291", + "issue:36032|issue:43742", + "issue:42915|issue:44589", + "issue:40444|issue:46326", + "issue:44062|issue:44987", + "issue:43531|issue:44877", + "issue:43611|issue:44534", + "issue:45812|issue:46491", + "issue:43425|issue:44403", + "issue:43296|issue:43531" + ], + "prepared_review_unit_hash": "9aa1dc935ab9cc153569fe90d0c8bee70063c940b0474d055e9595767269f25d", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36517, + "estimated_input_tokens": 9130, + "estimated_eval_tokens": 18516 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most items are distinct bugs/regressions across different models or subsystems. The only clear duplicate-like pair is the Qwen3.5 save_pretrained regression, where the later report narrows the same serialization problem to visual encoder keys.", + "confidence": 0.87, + "canonical_issue_reason": "issue:45216 is the broadest and earliest report of the Qwen3.5 save_pretrained regression, with a clear version boundary and the core serialization failure; issue:45357 reads as a later, narrower follow-up on the same bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45216 best captures the shared underlying bug because it describes the save_pretrained regression at the right abstraction level, while 45357 is a specific symptom of that same failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44849", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen model families and different bugs: hidden-states output logic vs RoPE position scaling for still images." + }, + { + "left": "issue:44464", + "right": "issue:45987", + "accept": false, + "reason": "Unrelated failures: chunked generation inconsistency with compiled forward vs StaticCache length return-type contract." + }, + { + "left": "issue:43262", + "right": "issue:43377", + "accept": false, + "reason": "Both audio-related, but one is a chat-template sampling-rate default and the other is a batching/padding-mask correctness bug in MIMI." + }, + { + "left": "issue:44987", + "right": "issue:44991", + "accept": false, + "reason": "Both are model-loading regressions, but they affect different models and appear to stem from different code paths." + }, + { + "left": "issue:43646", + "right": "issue:44534", + "accept": false, + "reason": "Custom model initialization breakage in notebooks is a different issue from non-persistent buffers being filled with junk." + }, + { + "left": "issue:43883", + "right": "issue:44991", + "accept": false, + "reason": "Molmo tied-weights attribute error is unrelated to tokenizer loading for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43404", + "right": "issue:44466", + "accept": false, + "reason": "Both mention tied weights/lm_head, but one is a model-specific missing tie and the other is a save_pretrained serialization inconsistency." + }, + { + "left": "issue:44403", + "right": "issue:45081", + "accept": false, + "reason": "Generic loading noise is not the same as the Mistral regex patch crash." + }, + { + "left": "issue:42915", + "right": "issue:45084", + "accept": false, + "reason": "Qwen3Moe FP8 config failure and non-template-node compile error are different concrete problems." + }, + { + "left": "issue:43994", + "right": "issue:44610", + "accept": false, + "reason": "Different vision issues: nonsensical SigLIP2 outputs vs wrong processor image size for OmDet-Turbo." + }, + { + "left": "issue:43377", + "right": "issue:43819", + "accept": false, + "reason": "Different audio codec/model bugs: padding-mask batching mismatch vs DAC latent/STE mismatch." + }, + { + "left": "issue:43322", + "right": "issue:43540", + "accept": false, + "reason": "Different multimodal models and failure modes: Llava Next loading segfault vs Qwen3OmniMoe video-processing ValueError." + }, + { + "left": "issue:28530", + "right": "issue:34567", + "accept": false, + "reason": "Trainer early-stopping configuration issue and token-count tracking bug are unrelated." + }, + { + "left": "issue:43645", + "right": "issue:44534", + "accept": false, + "reason": "Notebook custom-model initialization regression is separate from the non-persistent-buffer serialization bug." + }, + { + "left": "issue:43089", + "right": "issue:45987", + "accept": false, + "reason": "Generation performance overhead and StaticCache return-type mismatch are not the same bug." + }, + { + "left": "issue:44451", + "right": "issue:45081", + "accept": false, + "reason": "ScandiBERT loading failure is unrelated to the Mistral regex patch crash." + }, + { + "left": "issue:44987", + "right": "issue:45081", + "accept": false, + "reason": "Different models and different failure points; no shared underlying bug is evident." + }, + { + "left": "issue:44265", + "right": "issue:44610", + "accept": false, + "reason": "torch.export/torch_compilable_check failure is unrelated to an image-size processor bug." + }, + { + "left": "issue:44220", + "right": "issue:44683", + "accept": false, + "reason": "Feature-extraction helper bug and compiled flex_attention failure are unrelated." + }, + { + "left": "issue:44991", + "right": "issue:46315", + "accept": false, + "reason": "Tokenizer loading failure and slow convert_tokens_to_ids performance are different issues." + }, + { + "left": "issue:45216", + "right": "issue:45357", + "accept": true, + "reason": "Same Qwen3.5 save_pretrained regression: 45357 is a narrower follow-up describing incorrect visual encoder keys, which fits the same underlying serialization bug." + }, + { + "left": "issue:42915", + "right": "issue:44291", + "accept": false, + "reason": "Different incompatibilities: Qwen3Moe FP8 config failure vs init_empty_weights receiving an unexpected argument." + }, + { + "left": "issue:36032", + "right": "issue:43742", + "accept": false, + "reason": "Tokenizer method-name conflict and MobileLLM key error are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44589", + "accept": false, + "reason": "Different FP8-related failures with different symptoms and likely different code paths." + }, + { + "left": "issue:40444", + "right": "issue:46326", + "accept": false, + "reason": "IterableDataset multi-image finetuning failure is unrelated to mistral_common continue_final_message behavior." + }, + { + "left": "issue:44062", + "right": "issue:44987", + "accept": false, + "reason": "AddedToken constructor argument conflict and a model-loading regression for physical-intelligence/fast are unrelated." + }, + { + "left": "issue:43531", + "right": "issue:44877", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and granite_speech strict-config loading are different problems." + }, + { + "left": "issue:43611", + "right": "issue:44534", + "accept": false, + "reason": "base_model_prefix loading breakage is separate from junk-filled non-persistent buffers." + }, + { + "left": "issue:45812", + "right": "issue:46491", + "accept": false, + "reason": "Both tokenizer regressions, but for different model families and different observable failures." + }, + { + "left": "issue:43425", + "right": "issue:44403", + "accept": false, + "reason": "Torch version compatibility and generic transformer-loading noise do not describe the same bug." + }, + { + "left": "issue:43296", + "right": "issue:43531", + "accept": false, + "reason": "PaddleOCR-VL/vLLM loading failure is unrelated to the Qwen3-MoE sliding_window issue." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is grounded in the packet and stays conservative. The only accepted soft pair, issue:45216 vs issue:45357, is plausibly the same Qwen3.5 save_pretrained regression, with 45357 reading as a later/narrower follow-up. The rest of the pair verdicts are appropriately cautious and the summary does not overclaim broader duplication beyond what the titles support." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 11, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:33666", + "issue:36032", + "issue:40444", + "issue:40990", + "issue:41553", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:43317", + "issue:43377", + "issue:43381", + "issue:43421", + "issue:43482", + "issue:43525", + "issue:43531", + "issue:43582", + "issue:43632", + "issue:43746", + "issue:43792", + "issue:43818", + "issue:43856", + "issue:44079", + "issue:44291", + "issue:44403", + "issue:44462", + "issue:44479", + "issue:44534", + "issue:44568", + "issue:44589", + "issue:44683", + "issue:44756", + "issue:44811", + "issue:44829", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45081", + "issue:45325", + "issue:45357", + "issue:45479", + "issue:45491", + "issue:45538", + "issue:45542", + "issue:45663", + "issue:45812", + "issue:45923", + "issue:46326", + "issue:46402" + ], + "soft_pairs": [ + "issue:43632|issue:44756", + "issue:36032|issue:41720", + "issue:44479|issue:45325", + "issue:44977|issue:46326", + "issue:43792|issue:44589", + "issue:44403|issue:44991", + "issue:40990|issue:44811", + "issue:33666|issue:40444", + "issue:44683|issue:45020", + "issue:43381|issue:45923", + "issue:43632|issue:44534", + "issue:43421|issue:44568", + "issue:44291|issue:44589", + "issue:44829|issue:45479", + "issue:44079|issue:45663", + "issue:43482|issue:45538", + "issue:40444|issue:45325", + "issue:43482|issue:43525", + "issue:41553|issue:43582", + "issue:43317|issue:43856", + "issue:42175|issue:45542", + "issue:43818|issue:46402", + "issue:42796|issue:43746", + "issue:44462|issue:45812", + "issue:43818|issue:45357", + "issue:43377|issue:45491", + "issue:41720|issue:45081", + "issue:43531|issue:44291" + ], + "prepared_review_unit_hash": "bc30854faca267c4db5aba5a3acc1b040e1e7aeaec6c33ab3c301132643c77a0", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36295, + "estimated_input_tokens": 9074, + "estimated_eval_tokens": 18404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft pair is strong enough to treat as a confirmed duplicate. The Qwen2.5-VL pair (44479/45325) is thematically related, but the packet marks it deterministic_accept=false and does not establish that 45325 is the root cause of 44479; it should be treated as a tentative topical relation only.", + "confidence": 0.83, + "canonical_issue_reason": "issue:44479 is the best umbrella issue for the Qwen video-regression topic because it describes the broader user-facing failure across multiple Qwen VL variants. However, its relationship to issue:45325 is only tentative in the packet, not a confirmed duplicate/root-cause pairing.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the strongest triage anchor among the cluster because it covers the widest visible impact in the shared Qwen VL video-position-id regression theme, even though the evidence is not sufficient to collapse it with 45325.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43632", + "right": "issue:44756", + "accept": false, + "reason": "Different bugs: _is_hf_initialized / version regression vs mmap OOM on Strix Halo." + }, + { + "left": "issue:36032", + "right": "issue:41720", + "accept": false, + "reason": "Tokenizer attribute conflict and Qwen3 auto device-map CUDA assert are unrelated." + }, + { + "left": "issue:44479", + "right": "issue:45325", + "accept": false, + "reason": "Related Qwen2.5-VL video-position-id symptoms, but the packet does not conservatively prove the same underlying bug; deterministic_accept is false." + }, + { + "left": "issue:44977", + "right": "issue:46326", + "accept": false, + "reason": "Both are generation/tokenizer-adjacent, but the concrete failures and code paths differ." + }, + { + "left": "issue:43792", + "right": "issue:44589", + "accept": false, + "reason": "Whisper runtime failure and Float8 storage lookup are unrelated." + }, + { + "left": "issue:44403", + "right": "issue:44991", + "accept": false, + "reason": "Generic loading noise vs a specific tokenizer-loading regression for a model repo." + }, + { + "left": "issue:40990", + "right": "issue:44811", + "accept": false, + "reason": "Perplexity on gpt-oss-20b is unrelated to Whisper batch_decode skip_special_tokens behavior." + }, + { + "left": "issue:33666", + "right": "issue:40444", + "accept": false, + "reason": "Both are Qwen-VL related, but the training setups and failure modes are different." + }, + { + "left": "issue:44683", + "right": "issue:45020", + "accept": false, + "reason": "Torch flex_attention compilation and remote_code loading regressions are separate problems." + }, + { + "left": "issue:43381", + "right": "issue:45923", + "accept": false, + "reason": "Eval-mode gradient checkpointing limitation is not the same as a missing supports_gradient_checkpointing flag." + }, + { + "left": "issue:43632", + "right": "issue:44534", + "accept": false, + "reason": "Different internal regressions with different failure surfaces." + }, + { + "left": "issue:43421", + "right": "issue:44568", + "accept": false, + "reason": "Both involve special tokens, but one is runtime post-processor updates and the other is BOS/EOS insertion in a specific tokenizer." + }, + { + "left": "issue:44291", + "right": "issue:44589", + "accept": false, + "reason": "Unexpected _is_hf_initialized argument and missing Float8 storage object are unrelated loading errors." + }, + { + "left": "issue:44829", + "right": "issue:45479", + "accept": false, + "reason": "Both mention degenerate training, but the root causes differ (flash_attention_3 vs classification-loss setup)." + }, + { + "left": "issue:44079", + "right": "issue:45663", + "accept": false, + "reason": "ModelOutput key assignment and Gemma-4 FSDP2 shared_kv_states rebuilding are unrelated." + }, + { + "left": "issue:43482", + "right": "issue:45538", + "accept": false, + "reason": "GGUF loading failure and CLIPTokenizer model_max_length are separate tokenizer/model-loading issues." + }, + { + "left": "issue:40444", + "right": "issue:45325", + "accept": false, + "reason": "Both are Qwen2.5-VL-related, but one is multi-image finetuning and the other is temporal position-id scaling." + }, + { + "left": "issue:43482", + "right": "issue:43525", + "accept": false, + "reason": "GGUF loading failure is unrelated to Llama4Config missing pad_token_id." + }, + { + "left": "issue:41553", + "right": "issue:43582", + "accept": false, + "reason": "AutoTokenizer error messaging for Voxtral is unrelated to Apple Silicon allocator warmup TypeError." + }, + { + "left": "issue:43317", + "right": "issue:43856", + "accept": false, + "reason": "GPU/CPU offload loading failure and Qwen3 MoE memory usage are different bugs." + }, + { + "left": "issue:42175", + "right": "issue:45542", + "accept": false, + "reason": "Both touch TensorFlow packaging, but they describe different install states and failures." + }, + { + "left": "issue:43818", + "right": "issue:46402", + "accept": false, + "reason": "Video-LLaVA tower architecture issue and Sam3Model text-encoder loading are unrelated." + }, + { + "left": "issue:42796", + "right": "issue:43746", + "accept": false, + "reason": "Tokenizer BPE-marker handling and GraniteSpeech PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:44462", + "right": "issue:45812", + "accept": false, + "reason": "Both involve AutoTokenizer, but the repo-tokenizer precedence bug and Granite token-ID regression are not the same code-path." + }, + { + "left": "issue:43818", + "right": "issue:45357", + "accept": false, + "reason": "Video-LLaVA tower issue and Qwen3.5 save_pretrained visual-key regression are different regressions." + }, + { + "left": "issue:43377", + "right": "issue:45491", + "accept": false, + "reason": "MIMI padding-mask batching and Gemma3 sliding-window NaNs are different model bugs." + }, + { + "left": "issue:41720", + "right": "issue:45081", + "accept": false, + "reason": "CUDA assert during auto device mapping is unrelated to the Mistral tokenizer regex crash." + }, + { + "left": "issue:43531", + "right": "issue:44291", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and init_empty_weights/_is_hf_initialized loading errors are separate issues." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The summary overstates the evidence. In the packet, the only Qwen pair (issue:44479 / issue:45325) is explicitly marked deterministic_accept=false, so treating it as a confirmed duplicate/canonical pair is not conservative. The titles do show related Qwen2.5-VL/video-position-id behavior, but the claim that 45325 is the root cause behind 44479 is not grounded by the packet. The other soft-edge rejections are generally conservative enough, but the accepted pair and the canonical-issue rationale should be downgraded to a tentative thematic relation at most." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 12, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:35707", + "issue:39692", + "issue:40444", + "issue:41720", + "issue:42175", + "issue:42796", + "issue:42915", + "issue:43262", + "issue:43296", + "issue:43377", + "issue:43454", + "issue:43482", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43632", + "issue:43644", + "issue:43646", + "issue:43701", + "issue:43746", + "issue:43761", + "issue:43792", + "issue:43819", + "issue:43881", + "issue:43957", + "issue:43994", + "issue:44079", + "issue:44186", + "issue:44265", + "issue:44291", + "issue:44466", + "issue:44479", + "issue:44488", + "issue:44534", + "issue:44568", + "issue:44610", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:45072", + "issue:45325", + "issue:45381", + "issue:45800", + "issue:45823", + "issue:45834", + "issue:45920", + "issue:46305", + "issue:46315" + ], + "soft_pairs": [ + "issue:43611|issue:43646", + "issue:44265|issue:45072", + "issue:42175|issue:43296", + "issue:43994|issue:44898", + "issue:43792|issue:43957", + "issue:45325|issue:45834", + "issue:43454|issue:44466", + "issue:42915|issue:43482", + "issue:44079|issue:44534", + "issue:43632|issue:45072", + "issue:45325|issue:45381", + "issue:40444|issue:45381", + "issue:43644|issue:44079", + "issue:44291|issue:44877", + "issue:44479|issue:45823", + "issue:35707|issue:40444", + "issue:43761|issue:44079", + "issue:42796|issue:45381", + "issue:43296|issue:45800", + "issue:43606|issue:44265", + "issue:43701|issue:45325", + "issue:43881|issue:44488", + "issue:43606|issue:43819", + "issue:43454|issue:43550", + "issue:39692|issue:40444", + "issue:41720|issue:45823", + "issue:44186|issue:44265", + "issue:46305|issue:46315", + "issue:43262|issue:43746", + "issue:44568|issue:45920", + "issue:42796|issue:44977", + "issue:43377|issue:44610" + ], + "prepared_review_unit_hash": "f2f711e03ef1105d6f17c4e15667008c8acb30a7b9fac4dab547f8e0b0256db5", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37461, + "estimated_input_tokens": 9366, + "estimated_eval_tokens": 18988 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly heterogeneous, model-specific regressions. No soft pair looks like a clear duplicate; the closest ones are related by theme but not the same concrete bug. The best global issue anchor is the broad AutoTokenizer regression (issue 45920), since it is active, has inbound references, and covers a wide cross-model tokenizer failure mode.", + "confidence": 0.84, + "canonical_issue_reason": "Issue 45920 is the broadest and most central open regression here: it affects many models, has inbound references, and describes a reusable tokenizer failure pattern rather than a single model quirk.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45920 is the best cluster anchor because it is the most general, highest-impact issue in the set and is already being referenced externally.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43611", + "right": "issue:43646", + "accept": false, + "reason": "Both are v5 loading regressions, but one is about `base_model_prefix` and the other about custom model initialization; different failure points and likely different fixes." + }, + { + "left": "issue:44265", + "right": "issue:45072", + "accept": false, + "reason": "Unrelated bugs: `torch.export`/`torch_compilable_check` vs a bfloat16 dtype mismatch in inference." + }, + { + "left": "issue:42175", + "right": "issue:43296", + "accept": false, + "reason": "Different subsystems and symptoms: packaging/backend selection vs loading a specific model in vLLM." + }, + { + "left": "issue:43994", + "right": "issue:44898", + "accept": false, + "reason": "Both are vision-model regressions, but they target different models and different behaviors (nonsensical outputs vs image-size handling)." + }, + { + "left": "issue:43792", + "right": "issue:43957", + "accept": false, + "reason": "One is a model-specific runtime failure, the other is a meta-device loading regression; not the same code-path bug." + }, + { + "left": "issue:45325", + "right": "issue:45834", + "accept": false, + "reason": "Both involve multimodal position/index handling, but one is a Qwen2.5-VL rope-index scaling bug and the other is a Kosmos2.5 OCR length/index error." + }, + { + "left": "issue:43454", + "right": "issue:44466", + "accept": false, + "reason": "Both mention `lm_head`, but one is about missing weight tying and the other about inconsistent serialization; related area, not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:43482", + "accept": false, + "reason": "Different models and different loading/quantization paths; no shared concrete defect." + }, + { + "left": "issue:44079", + "right": "issue:44534", + "accept": false, + "reason": "`ModelOutput` key assignment and non-persistent buffer initialization are unrelated." + }, + { + "left": "issue:43632", + "right": "issue:45072", + "accept": false, + "reason": "Different regressions: `_is_hf_initialized` handling vs dtype mismatches in inference." + }, + { + "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 `vision_position_ids`; close but not clearly the same bug." + }, + { + "left": "issue:40444", + "right": "issue:45381", + "accept": false, + "reason": "Both concern Qwen2.5-VL multimodal inputs, but one is finetuning with multiple images per prompt and the other is a video position-id regression." + }, + { + "left": "issue:43644", + "right": "issue:44079", + "accept": false, + "reason": "One is buffer initialization junk, the other is `ModelOutput` field assignment; no duplicate-level overlap." + }, + { + "left": "issue:44291", + "right": "issue:44877", + "accept": false, + "reason": "Both are loading-related, but one is an `init_empty_weights` argument regression and the other is strict config validation for GraniteSpeech." + }, + { + "left": "issue:44479", + "right": "issue:45823", + "accept": false, + "reason": "Both involve device/runtime issues, but they affect different models and different mechanisms (`device_map`/video regression vs Gemma4 PLE forward mismatch)." + }, + { + "left": "issue:35707", + "right": "issue:40444", + "accept": false, + "reason": "Generation with `inputs_embeds`/`past_key_values` is unrelated to Qwen2.5-VL multi-image finetuning." + }, + { + "left": "issue:43761", + "right": "issue:44079", + "accept": false, + "reason": "Different failures: CLIP hidden states not returned vs `ModelOutput` key bookkeeping." + }, + { + "left": "issue:42796", + "right": "issue:45381", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to Qwen2.5-VL video position IDs." + }, + { + "left": "issue:43296", + "right": "issue:45800", + "accept": false, + "reason": "A model-loading failure in vLLM is not the same as a broad torch/transformers version incompatibility report." + }, + { + "left": "issue:43606", + "right": "issue:44265", + "accept": false, + "reason": "CPU offload device mismatch and `torch.export` failure from `torch_compilable_check` are different bugs." + }, + { + "left": "issue:43701", + "right": "issue:45325", + "accept": false, + "reason": "Checkpoint key mismatch and Qwen2.5-VL rope-index scaling are unrelated." + }, + { + "left": "issue:43881", + "right": "issue:44488", + "accept": false, + "reason": "Different model-loading problems with no shared code-path evidence." + }, + { + "left": "issue:43606", + "right": "issue:43819", + "accept": false, + "reason": "Device mismatch on CPU offload is unrelated to DAC latent reconstruction mismatch." + }, + { + "left": "issue:43454", + "right": "issue:43550", + "accept": false, + "reason": "Weight-tying/generation correctness and torch.compile SDPA issues are different defects." + }, + { + "left": "issue:39692", + "right": "issue:40444", + "accept": false, + "reason": "Documentation/example errors are not the same bug as a finetuning failure with multiple images per prompt." + }, + { + "left": "issue:41720", + "right": "issue:45823", + "accept": false, + "reason": "Both mention device mapping, but the affected models and concrete failures differ substantially." + }, + { + "left": "issue:44186", + "right": "issue:44265", + "accept": false, + "reason": "Tokenizer padding/truncation crashes and `torch.export`/compilability issues are unrelated." + }, + { + "left": "issue:46305", + "right": "issue:46315", + "accept": false, + "reason": "Sliding-window masking off-by-one and tokenizer slowdown are entirely different regressions." + }, + { + "left": "issue:43262", + "right": "issue:43746", + "accept": false, + "reason": "Audio chat-template sample-rate defaults and GraniteSpeech PEFT checkpoint loading are separate issues." + }, + { + "left": "issue:44568", + "right": "issue:45920", + "accept": false, + "reason": "Both are tokenizer-related v5 regressions, but one is special-token insertion for a specific tokenizer and the other is widespread wrong token IDs across many models." + }, + { + "left": "issue:42796", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer BPE-marker handling and Qwen3.5 flash-attention generation are unrelated." + }, + { + "left": "issue:43377", + "right": "issue:44610", + "accept": false, + "reason": "Padding-mask behavior in MIMI encoder and OmDet-Turbo processor size mismatch are different model pipeline 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": 20, + "review_unit_index": 13, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:26413", + "issue:28530", + "issue:34634", + "issue:35707", + "issue:36331", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42796", + "issue:42915", + "issue:42947", + "issue:43065", + "issue:43257", + "issue:43317", + "issue:43377", + "issue:43404", + "issue:43450", + "issue:43454", + "issue:43526", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43708", + "issue:43720", + "issue:43854", + "issue:43873", + "issue:43881", + "issue:43883", + "issue:44008", + "issue:44186", + "issue:44387", + "issue:44423", + "issue:44466", + "issue:44756", + "issue:44877", + "issue:44898", + "issue:44977", + "issue:44991", + "issue:45072", + "issue:45357", + "issue:45406", + "issue:45584", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45823", + "issue:45923", + "issue:46402" + ], + "soft_pairs": [ + "issue:40990|issue:45584", + "issue:43454|issue:45072", + "issue:43317|issue:43873", + "issue:43257|issue:45923", + "issue:43377|issue:44898", + "issue:40990|issue:42947", + "issue:43572|issue:45759", + "issue:43873|issue:45701", + "issue:43454|issue:43720", + "issue:42796|issue:43854", + "issue:43526|issue:43883", + "issue:44466|issue:45357", + "issue:41720|issue:44977", + "issue:43881|issue:44991", + "issue:26413|issue:28530", + "issue:42617|issue:42915", + "issue:35707|issue:41720", + "issue:43065|issue:46402", + "issue:34634|issue:35707", + "issue:43450|issue:45701", + "issue:44877|issue:44991", + "issue:43404|issue:45357", + "issue:44186|issue:45072", + "issue:36331|issue:42915", + "issue:43550|issue:43720", + "issue:43577|issue:43708", + "issue:44008|issue:45823", + "issue:36331|issue:42617", + "issue:44423|issue:45406", + "issue:44387|issue:45704", + "issue:43873|issue:44756" + ], + "prepared_review_unit_hash": "6303265401f4824c1b9d3a909b88bc408bb8ab171260534512ce78630743781a", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37190, + "estimated_input_tokens": 9298, + "estimated_eval_tokens": 18852 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is highly heterogeneous and does not look like a duplicate set; the only recurring theme is a loose family of loading/offloading/quantization-related reports, with no pair meeting a strict same-bug threshold.", + "confidence": 0.33, + "canonical_issue_reason": "issue:43873 is the most central of the bunch because it is the broadest loading/offloading report and it appears in multiple soft-similarity links, but the cluster overall is too mixed to treat it as a true duplicate anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 is the best single issue to represent the cluster\u2019s loose offloading/quantization theme, but it is not a good global duplicate canonical because most other issues are unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:40990", + "right": "issue:45584", + "accept": false, + "reason": "Both are model behavior bugs, but one is perplexity on gpt-oss-20b and the other is Whisper empty-transcription handling; no shared code-path is evident." + }, + { + "left": "issue:43454", + "right": "issue:45072", + "accept": false, + "reason": "Different failures: tied lm_head weights in AyaVision vs bfloat16 dtype mismatch in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:43317", + "right": "issue:43873", + "accept": false, + "reason": "Both mention offloading/quantization, but one is dequantized model loading with device_map=auto and the other is a broader CUDA memory/offloading complaint; too generic to merge." + }, + { + "left": "issue:43257", + "right": "issue:45923", + "accept": false, + "reason": "Accelerate+DeepSpeed MOE weight conversion and a missing gradient-checkpointing flag are unrelated issues." + }, + { + "left": "issue:43377", + "right": "issue:44898", + "accept": false, + "reason": "Different processors and different bugs: padding-mask-dependent encoder outputs vs image classification with interpolate_pos_encoding." + }, + { + "left": "issue:40990", + "right": "issue:42947", + "accept": false, + "reason": "Perplexity/eval-quality issue vs gradient-checkpointing ineffectiveness with PEFT LoRA; not the same bug." + }, + { + "left": "issue:43572", + "right": "issue:45759", + "accept": false, + "reason": "Both concern config/model loading, but one is a missing pad token field and the other is unwrapping composite text_config; different code paths." + }, + { + "left": "issue:43873", + "right": "issue:45701", + "accept": false, + "reason": "Offloading/quantization problems are not the same as version-dependent tokenization changes." + }, + { + "left": "issue:43454", + "right": "issue:43720", + "accept": false, + "reason": "AyaVision tied-weight serialization and BitNet packed-weight unpacking are distinct model-loading bugs." + }, + { + "left": "issue:42796", + "right": "issue:43854", + "accept": false, + "reason": "Tokenizer BPE-marker handling is unrelated to GLM-4.7 model loading failures." + }, + { + "left": "issue:43526", + "right": "issue:43883", + "accept": false, + "reason": "BEiT label reduction and missing all_tied_weights_keys are unrelated implementation bugs." + }, + { + "left": "issue:44466", + "right": "issue:45357", + "accept": false, + "reason": "Both involve save/load behavior, but one is tied lm_head serialization and the other is incorrect visual encoder keys; different model components." + }, + { + "left": "issue:41720", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen-family generation issues, but one is an auto-device-map CUDA assert and the other is a flash-attention generation bug; not clearly the same defect." + }, + { + "left": "issue:43881", + "right": "issue:44991", + "accept": false, + "reason": "Different failure modes in model/tokenizer loading for unrelated models." + }, + { + "left": "issue:26413", + "right": "issue:28530", + "accept": false, + "reason": "Checkpoint resume/epoch iterator failure is unrelated to early stopping metric selection." + }, + { + "left": "issue:42617", + "right": "issue:42915", + "accept": false, + "reason": "3D parallel execution failure and Qwen3Moe fine-grained FP8 loading failure are different problems." + }, + { + "left": "issue:35707", + "right": "issue:41720", + "accept": false, + "reason": "Progressive generation with inputs_embeds/past_key_values is unrelated to Qwen3 auto device-map CUDA assertion." + }, + { + "left": "issue:43065", + "right": "issue:46402", + "accept": false, + "reason": "A dummy Conv2d in Sam3PixelDecoder and missing Sam3Model text-encoder weights are different Sam3 bugs." + }, + { + "left": "issue:34634", + "right": "issue:35707", + "accept": false, + "reason": "BarkProcessor voice_preset handling has no clear overlap with progressive generation using inputs_embeds." + }, + { + "left": "issue:43450", + "right": "issue:45701", + "accept": false, + "reason": "Video processor batch-shape bug and tokenization version-change bug are unrelated subsystems." + }, + { + "left": "issue:44877", + "right": "issue:44991", + "accept": false, + "reason": "Strict config rejection for granite_speech is not the same as tokenizer loading failure for EMBEDDIA/est-roberta." + }, + { + "left": "issue:43404", + "right": "issue:45357", + "accept": false, + "reason": "Different tied-weight/save-pretrained issues affecting different model architectures." + }, + { + "left": "issue:44186", + "right": "issue:45072", + "accept": false, + "reason": "Tokenizer crash on NER/batched padding is unrelated to dtype mismatch in inference." + }, + { + "left": "issue:36331", + "right": "issue:42915", + "accept": false, + "reason": "CustomTrainer API breakage and Qwen3Moe FP8 loading failure are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:43720", + "accept": false, + "reason": "torch.compile+SDPA failure in Bamba is not the same as accelerate loading not unpacking BitNet weights." + }, + { + "left": "issue:43577", + "right": "issue:43708", + "accept": false, + "reason": "Blip2 dtype not propagating and resume_from_checkpoint max_steps miscalculation are unrelated bugs." + }, + { + "left": "issue:44008", + "right": "issue:45823", + "accept": false, + "reason": "Gemma 3n variable-name collision in forward is unrelated to Gemma4 PLE device placement mismatch." + }, + { + "left": "issue:36331", + "right": "issue:42617", + "accept": false, + "reason": "Trainer compute_loss API breakage and 3D parallel execution failure do not share a concrete code path." + }, + { + "left": "issue:44423", + "right": "issue:45406", + "accept": false, + "reason": "Both affect serve, but the first is multimodal continuous-batching with a string/.to error and the second is a missing _tokenizer on Gemma4Processor; different root causes." + }, + { + "left": "issue:44387", + "right": "issue:45704", + "accept": false, + "reason": "Int4 CUDA reserved-memory growth and an apex RMSNorm memory leak are different memory issues with different causes." + }, + { + "left": "issue:43873", + "right": "issue:44756", + "accept": false, + "reason": "Quantization/offloading behavior is too broad a commonality; disabling mmap on Strix Halo is a platform-specific OOM workaround, 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": 20, + "review_unit_index": 14, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28530", + "issue:29159", + "issue:30064", + "issue:33666", + "issue:34634", + "issue:36010", + "issue:36032", + "issue:40990", + "issue:41720", + "issue:42617", + "issue:42915", + "issue:43295", + "issue:43316", + "issue:43317", + "issue:43322", + "issue:43421", + "issue:43526", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43606", + "issue:43653", + "issue:43720", + "issue:43749", + "issue:43782", + "issue:43881", + "issue:44186", + "issue:44361", + "issue:44492", + "issue:44509", + "issue:44534", + "issue:44610", + "issue:44671", + "issue:44743", + "issue:44898", + "issue:44913", + "issue:45161", + "issue:45357", + "issue:45406", + "issue:45464", + "issue:45632", + "issue:45684", + "issue:45701", + "issue:45759", + "issue:46050", + "issue:46291", + "issue:46402" + ], + "soft_pairs": [ + "issue:43749|issue:44534", + "issue:43550|issue:43606", + "issue:44610|issue:44898", + "issue:44186|issue:44610", + "issue:43295|issue:45701", + "issue:45632|issue:45684", + "issue:42915|issue:43317", + "issue:43322|issue:46050", + "issue:41720|issue:43322", + "issue:41720|issue:42617", + "issue:40990|issue:45701", + "issue:36010|issue:42617", + "issue:41720|issue:45161", + "issue:43531|issue:43572", + "issue:42915|issue:46291", + "issue:28530|issue:33666", + "issue:43526|issue:43881", + "issue:44671|issue:44743", + "issue:44186|issue:44361", + "issue:43540|issue:43782", + "issue:43316|issue:44913", + "issue:28530|issue:30064", + "issue:29159|issue:36032", + "issue:43550|issue:43653", + "issue:43606|issue:43720", + "issue:30064|issue:34634", + "issue:45406|issue:45464", + "issue:43421|issue:45701", + "issue:42915|issue:43322", + "issue:45357|issue:46402", + "issue:44492|issue:44509", + "issue:43606|issue:43653", + "issue:43531|issue:45759" + ], + "prepared_review_unit_hash": "dc0c2b7e41e8670db7e921d9d2edf9f4ddfe8cff7da5857c7eebfeed8719bc47", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 33, + "serialized_chars": 37345, + "estimated_input_tokens": 9337, + "estimated_eval_tokens": 18930 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly semantic noise: the issues span unrelated models, tokenizer behaviors, loading paths, docs, and runtime errors. None of the soft pairs look like the same underlying bug/change, so no duplicate consolidation is warranted.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43749", + "right": "issue:44534", + "accept": false, + "reason": "One is FSDP CPU/RAM-efficient loading breaking; the other is v5 non-persistent buffer initialization. Different failure modes and code paths." + }, + { + "left": "issue:43550", + "right": "issue:43606", + "accept": false, + "reason": "Different models and bugs: torch.compile+SDPA on Bamba vs CPU offload device mismatch on Bark." + }, + { + "left": "issue:44610", + "right": "issue:44898", + "accept": false, + "reason": "Both involve image resolution, but they are separate model-specific preprocessing bugs with different expected sizes and failure points." + }, + { + "left": "issue:44186", + "right": "issue:44610", + "accept": false, + "reason": "LayoutLMv2 tokenizer NER/padding crash is unrelated to OmDet-Turbo image-size mismatch." + }, + { + "left": "issue:43295", + "right": "issue:45701", + "accept": false, + "reason": "The first is a regression in processor/tokenizer access; the second is a broader tokenization-change report. Not the same concrete bug." + }, + { + "left": "issue:45632", + "right": "issue:45684", + "accept": false, + "reason": "Cache-path collisions for trust_remote_code and read-only permission propagation in save_pretrained are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:43317", + "accept": false, + "reason": "Both are loading/quantization-related, but they describe different models, configs, and distinct offload failures." + }, + { + "left": "issue:43322", + "right": "issue:46050", + "accept": false, + "reason": "Segfault when loading Llava Next is not the same as sinq quantization save/load breakage." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Different model families and symptoms: cudaErrorAssert on Qwen3 vs segfault on Llava Next." + }, + { + "left": "issue:41720", + "right": "issue:42617", + "accept": false, + "reason": "Qwen3 auto device mapping failure is unrelated to not being able to run 3d_parallel.py." + }, + { + "left": "issue:40990", + "right": "issue:45701", + "accept": false, + "reason": "High perplexity on a specific model/dataset is not the same as a general tokenization-version change report." + }, + { + "left": "issue:36010", + "right": "issue:42617", + "accept": false, + "reason": "GenerationMixin import failure and a 3d_parallel.py runtime issue are unrelated." + }, + { + "left": "issue:41720", + "right": "issue:45161", + "accept": false, + "reason": "Different Qwen/GPT-OSS-related loading issues with different symptoms and root causes." + }, + { + "left": "issue:43531", + "right": "issue:43572", + "accept": false, + "reason": "Sliding-window behavior in Qwen3-MoE is unrelated to StableLmConfig missing pad_token_idx." + }, + { + "left": "issue:42915", + "right": "issue:46291", + "accept": false, + "reason": "FineGrainedFP8Config failure and pipelines import dependency-version failure are different problems." + }, + { + "left": "issue:28530", + "right": "issue:33666", + "accept": false, + "reason": "Early stopping metric lookup and Qwen2-VL multi-GPU training are unrelated." + }, + { + "left": "issue:43526", + "right": "issue:43881", + "accept": false, + "reason": "BEiT image label reduction bug is unrelated to glm-4v-9b loading failure." + }, + { + "left": "issue:44671", + "right": "issue:44743", + "accept": false, + "reason": "CamemBERT masked-LM prediction regression and Qwen3 modular recurrent-state reset are different code paths." + }, + { + "left": "issue:44186", + "right": "issue:44361", + "accept": false, + "reason": "Tokenizer crashes on NER/padding vs MLukeTokenizer task AttributeError: related subsystem, but not the same bug." + }, + { + "left": "issue:43540", + "right": "issue:43782", + "accept": false, + "reason": "Video-input processing error in Qwen3OmniMoe is different from weight_only load failure in Qwen3VL." + }, + { + "left": "issue:43316", + "right": "issue:44913", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and GPTNeoX rotary_pct reload issue are unrelated config bugs." + }, + { + "left": "issue:28530", + "right": "issue:30064", + "accept": false, + "reason": "Early stopping/eval metric handling has nothing to do with void segmentation map processing." + }, + { + "left": "issue:29159", + "right": "issue:36032", + "accept": false, + "reason": "Tokenizer inconsistency between slow/fast and T5 add_special_tokens conflict are distinct issues." + }, + { + "left": "issue:43550", + "right": "issue:43653", + "accept": false, + "reason": "Bamba torch.compile+SDPA failure and BigBird mask-token registration bug are unrelated." + }, + { + "left": "issue:43606", + "right": "issue:43720", + "accept": false, + "reason": "Both involve loading, but Bark CPU offload mismatch and BitNet packed-weight unpacking are different fixes." + }, + { + "left": "issue:30064", + "right": "issue:34634", + "accept": false, + "reason": "ImageProcessor void segmentation-map handling and BarkProcessor voice_preset are unrelated processor bugs." + }, + { + "left": "issue:45406", + "right": "issue:45464", + "accept": false, + "reason": "Gemma4Processor missing _tokenizer on serve is unrelated to Qwen3.5 streaming chat/completions failure." + }, + { + "left": "issue:43421", + "right": "issue:45701", + "accept": false, + "reason": "Runtime special-token/post-processor update behavior is not the same as a generic tokenization-change report." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Qwen3.5 visual-encoder save_pretrained keys and Sam3 text-encoder weights loading are different model-specific serialization bugs." + }, + { + "left": "issue:44492", + "right": "issue:44509", + "accept": false, + "reason": "A typo in cache strategies and outdated docs about removed pipelines are both docs-related, but not the same issue." + }, + { + "left": "issue:43606", + "right": "issue:43653", + "accept": false, + "reason": "Different tokenizer/model bugs: Bark CPU offload mismatch vs BigBird special-token registration." + }, + { + "left": "issue:43531", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3-MoE sliding-window behavior and AutoModelForCausalLM text_config unwrapping are separate Qwen-related issues." + }, + { + "left": "issue:41720", + "right": "issue:43322", + "accept": false, + "reason": "Qwen3 auto device mapping and Llava Next loading segfault are not the same failure." + }, + { + "left": "issue:42915", + "right": "issue:43322", + "accept": false, + "reason": "FineGrainedFP8Config failure and Llava Next segfault are unrelated model-loading bugs." + }, + { + "left": "issue:45357", + "right": "issue:46402", + "accept": false, + "reason": "Same general area of model serialization/loading, but different models, different attributes, and different root causes." + } + ] + }, + "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": 20, + "review_unit_index": 15, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:33666", + "issue:34567", + "issue:34634", + "issue:35707", + "issue:39692", + "issue:41720", + "issue:42490", + "issue:42886", + "issue:43329", + "issue:43381", + "issue:43388", + "issue:43421", + "issue:43525", + "issue:43531", + "issue:43653", + "issue:43746", + "issue:43844", + "issue:43873", + "issue:43927", + "issue:44062", + "issue:44186", + "issue:44368", + "issue:44451", + "issue:44514", + "issue:44545", + "issue:44589", + "issue:44610", + "issue:44843", + "issue:44898", + "issue:44918", + "issue:44936", + "issue:44987", + "issue:44991", + "issue:45042", + "issue:45059", + "issue:45072", + "issue:45081", + "issue:45356", + "issue:45357", + "issue:45588", + "issue:45663", + "issue:45704", + "issue:45759", + "issue:45859", + "issue:45923", + "issue:46032", + "issue:46463" + ], + "soft_pairs": [ + "issue:34567|issue:35707", + "issue:43388|issue:45059", + "issue:43873|issue:45704", + "issue:42490|issue:43329", + "issue:44451|issue:44991", + "issue:43531|issue:44589", + "issue:33666|issue:35707", + "issue:44898|issue:45072", + "issue:44514|issue:44545", + "issue:45759|issue:45859", + "issue:41720|issue:43746", + "issue:45588|issue:45759", + "issue:44368|issue:45357", + "issue:44062|issue:44451", + "issue:43421|issue:43927", + "issue:44987|issue:45042", + "issue:39692|issue:44610", + "issue:43381|issue:43844", + "issue:42886|issue:44843", + "issue:45704|issue:45923", + "issue:44918|issue:44936", + "issue:46032|issue:46463", + "issue:43421|issue:43653", + "issue:45704|issue:45859", + "issue:45663|issue:45923", + "issue:34634|issue:41720", + "issue:43525|issue:43531", + "issue:43421|issue:44186", + "issue:43381|issue:45704", + "issue:39692|issue:43873", + "issue:45081|issue:45356" + ], + "prepared_review_unit_hash": "1ad4db2b9f3557b752ce8c6e628ac0b7df5b743cf7f00b808fd497937fd6196e", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 36897, + "estimated_input_tokens": 9225, + "estimated_eval_tokens": 18706 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly heterogeneous: training, tokenizer, multimodal processor, offload, and CI bugs that do not share a single underlying fix. The only clear duplicate pair is the two Qwen2.5-VL `apply_chat_template(..., padding=False)` reports.", + "confidence": 0.93, + "canonical_issue_reason": "Use issue:44514 as the canonical issue for the only true duplicate pair; issue:44545 is the same Qwen2_5_VLProcessor batched `apply_chat_template` crash with `padding=False`, just phrased slightly differently.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44514 is the best issue representative here because it is a crisp, reproducible bug report and has an exact duplicate (issue:44545).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:34567", + "right": "issue:35707", + "accept": false, + "reason": "Different failures: TrainerState token counting vs progressive generation with `inputs_embeds`/`past_key_values`; no shared code-path bug." + }, + { + "left": "issue:43388", + "right": "issue:45059", + "accept": false, + "reason": "Unrelated subsystems: metric gathering/label truncation vs SAM3 text+box prompt behavior." + }, + { + "left": "issue:43873", + "right": "issue:45704", + "accept": false, + "reason": "Quantization/offloading behavior vs T5 apex RMSNorm memory leak; different bug classes and fixes." + }, + { + "left": "issue:42490", + "right": "issue:43329", + "accept": false, + "reason": "Local-path slash import warning is unrelated to the multimodal token counting video-branch NameError." + }, + { + "left": "issue:44451", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer loading regressions, but for different models and different failure modes." + }, + { + "left": "issue:43531", + "right": "issue:44589", + "accept": false, + "reason": "Qwen3-MoE sliding_window issue is unrelated to the Float8 storage type error." + }, + { + "left": "issue:33666", + "right": "issue:35707", + "accept": false, + "reason": "Multi-GPU training request vs generation correctness bug; not the same underlying issue." + }, + { + "left": "issue:44898", + "right": "issue:45072", + "accept": false, + "reason": "Perceiver image-resolution failure and dtype mismatch in SwitchTransformers/TimmWrapperModel are different code paths." + }, + { + "left": "issue:44514", + "right": "issue:44545", + "accept": true, + "reason": "Same bug and same code path: `Qwen2_5_VLProcessor.apply_chat_template` crashes on batched input when `padding=False`." + }, + { + "left": "issue:45759", + "right": "issue:45859", + "accept": false, + "reason": "Composite multimodal config unwrapping vs CPU offload incompatibility in Qwen3.5 MoE rotary embedding; not mergeable as one fix." + }, + { + "left": "issue:41720", + "right": "issue:43746", + "accept": false, + "reason": "Qwen3 auto device mapping CUDA assert and GraniteSpeech local PEFT checkpoint loading are unrelated." + }, + { + "left": "issue:45588", + "right": "issue:45759", + "accept": false, + "reason": "Flash-attention `s_aux=None` crash is unrelated to Qwen multimodal config handling." + }, + { + "left": "issue:44368", + "right": "issue:45357", + "accept": false, + "reason": "Warning about `tie_word_embeddings` during LoRA fine-tuning is not the same as saving incorrect visual encoder keys." + }, + { + "left": "issue:44062", + "right": "issue:44451", + "accept": false, + "reason": "`AddedToken(..., special=...)` construction error is unrelated to the ScandiBERT loading failure." + }, + { + "left": "issue:43421", + "right": "issue:43531", + "accept": false, + "reason": "Runtime special-token post-processor update issue is unrelated to Qwen3-MoE `sliding_window` behavior." + }, + { + "left": "issue:44987", + "right": "issue:45042", + "accept": false, + "reason": "Loading a specific model in transformers>=5.1.0 and PIL backend processors requiring torchvision are separate regressions." + }, + { + "left": "issue:39692", + "right": "issue:44610", + "accept": false, + "reason": "SigLIP2 docs example errors and OmDet-Turbo processor/model resolution mismatch are different issues." + }, + { + "left": "issue:43381", + "right": "issue:43844", + "accept": false, + "reason": "Gradient checkpointing eval-mode restriction is unrelated to DeepSpeed ZeRO-3 gradient growth." + }, + { + "left": "issue:42886", + "right": "issue:44843", + "accept": false, + "reason": "Tokenizer cache offline failure and unconditional `model_info()` call in `_patch_mistral_regex` are distinct offline-loading bugs." + }, + { + "left": "issue:45704", + "right": "issue:45923", + "accept": false, + "reason": "T5 apex RMSNorm leak is unrelated to missing `supports_gradient_checkpointing` on Nemotron trust_remote_code." + }, + { + "left": "issue:44918", + "right": "issue:44936", + "accept": false, + "reason": "SFT embedding unpacking failure is not the same as `trainer.evaluate()` failing after `trainer.train()`." + }, + { + "left": "issue:46032", + "right": "issue:46463", + "accept": false, + "reason": "Mamba2Mixer `use_cache` correctness bug and T5 decoder-input error message bug are unrelated." + }, + { + "left": "issue:43421", + "right": "issue:43653", + "accept": false, + "reason": "Tokenizer backend post-processor update bug is unrelated to BigBirdTokenizer special-token registration/decoding." + }, + { + "left": "issue:45704", + "right": "issue:45859", + "accept": false, + "reason": "Memory leak in T5 RMSNorm is unrelated to Qwen3.5 MoE CPU-offload incompatibility." + }, + { + "left": "issue:45663", + "right": "issue:45923", + "accept": false, + "reason": "Gemma-4 FSDP2 shared_kv_states bug is unrelated to a missing class flag on Nemotron." + }, + { + "left": "issue:34634", + "right": "issue:41720", + "accept": false, + "reason": "BarkProcessor voice preset parsing and Qwen3 auto device mapping CUDA assert are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43531", + "accept": false, + "reason": "Llama4Config missing `pad_token_id` and Qwen3-MoE `sliding_window` are different configuration bugs." + }, + { + "left": "issue:43421", + "right": "issue:44186", + "accept": false, + "reason": "TokenizersBackend special-token update bug is unrelated to LayoutLMv2 NER/padding crash." + }, + { + "left": "issue:43381", + "right": "issue:45704", + "accept": false, + "reason": "Gradient checkpointing eval-mode issue and T5 memory leak are separate problems." + }, + { + "left": "issue:39692", + "right": "issue:43873", + "accept": false, + "reason": "SigLIP2 documentation errors are unrelated to quantization/offloading behavior." + }, + { + "left": "issue:45081", + "right": "issue:45356", + "accept": false, + "reason": "Both involve tokenizer regex/codec handling, but the concrete failures differ: crash in `_patch_mistral_regex` vs Kimi-K2.5 regression and misleading warning." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The duplicate call on issue:44514 / issue:44545 is well supported by the titles, and the other soft-edge rejections do not overclaim sameness based on the provided evidence." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 16, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:30064", + "issue:36010", + "issue:36331", + "issue:42915", + "issue:43295", + "issue:43335", + "issue:43388", + "issue:43531", + "issue:43540", + "issue:43550", + "issue:43596", + "issue:43618", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43716", + "issue:43720", + "issue:43746", + "issue:43756", + "issue:43782", + "issue:43819", + "issue:43950", + "issue:43994", + "issue:44008", + "issue:44062", + "issue:44077", + "issue:44186", + "issue:44220", + "issue:44488", + "issue:44625", + "issue:44661", + "issue:44671", + "issue:44843", + "issue:44877", + "issue:44898", + "issue:45072", + "issue:45081", + "issue:45183", + "issue:45356", + "issue:45381", + "issue:45479", + "issue:45834", + "issue:45907", + "issue:45987", + "issue:46032", + "issue:46050", + "issue:46294" + ], + "soft_pairs": [ + "issue:43596|issue:44220", + "issue:45356|issue:45907", + "issue:43540|issue:46294", + "issue:44625|issue:45479", + "issue:43994|issue:44671", + "issue:43782|issue:45381", + "issue:30064|issue:36010", + "issue:43720|issue:43819", + "issue:43653|issue:43720", + "issue:44220|issue:44661", + "issue:44077|issue:44625", + "issue:43746|issue:46050", + "issue:43618|issue:43697", + "issue:42915|issue:43746", + "issue:43550|issue:45072", + "issue:43295|issue:45834", + "issue:43550|issue:43819", + "issue:43540|issue:44008", + "issue:44186|issue:44898", + "issue:44062|issue:44488", + "issue:43653|issue:43819", + "issue:45183|issue:45479", + "issue:43531|issue:43643", + "issue:43335|issue:43756", + "issue:43388|issue:46032", + "issue:43643|issue:44877", + "issue:43716|issue:44008", + "issue:44843|issue:45081", + "issue:30064|issue:36331", + "issue:43950|issue:45987" + ], + "prepared_review_unit_hash": "52dd0dd85f55d8e52129119895830228d235086b37f631b3e704dca1fb695684", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 36628, + "estimated_input_tokens": 9157, + "estimated_eval_tokens": 18570 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly noisy: the soft pairs are similar only at a broad topic level, not the same underlying bug/change. No PRs are present, and the closest thing to a theme is a small Mistral/tokenizer-regex subgroup.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45081 is the closest anchor in an otherwise heterogeneous set: it is a concrete, well-referenced tokenizer-regex regression, and several nearby items are at least tokenizer/loading related.", + "canonical_pr_reason": null, + "best_issue_reason": "If a representative issue must be chosen, issue:45081 is the strongest fit because it is specific, actionable, and has more discussion/inbound references than the other near-matches.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43596", + "right": "issue:44220", + "accept": false, + "reason": "Different subsystems and failure modes: DeepSpeed/BERT init index error vs audio feature extraction." + }, + { + "left": "issue:45356", + "right": "issue:45907", + "accept": false, + "reason": "Unrelated bugs: tokenizer codec/regex regression vs Hub list_repo_files revision forwarding." + }, + { + "left": "issue:43540", + "right": "issue:46294", + "accept": false, + "reason": "Both are video-related, but they target different Qwen models and different code paths/errors." + }, + { + "left": "issue:44625", + "right": "issue:45479", + "accept": false, + "reason": "One is Qwen3.5 config propagation; the other is a generic sequence-classification loss issue." + }, + { + "left": "issue:43994", + "right": "issue:44671", + "accept": false, + "reason": "Different models and symptoms: SigLIP2 pipeline outputs vs CamemBERT masked-LM predictions." + }, + { + "left": "issue:43782", + "right": "issue:45381", + "accept": false, + "reason": "Different models and bugs: Qwen3VL weight_only load failure vs Qwen2.5-VL video position-id mismatch." + }, + { + "left": "issue:30064", + "right": "issue:36010", + "accept": false, + "reason": "Completely unrelated: image processor void segmentation maps vs GenerationMixin import error." + }, + { + "left": "issue:43720", + "right": "issue:43819", + "accept": false, + "reason": "Different code paths: packed-weight loading in BitNet vs DAC latent/STE behavior." + }, + { + "left": "issue:43653", + "right": "issue:43720", + "accept": false, + "reason": "Tokenizer special-token registration bug vs BitNet packed-weight loading bug; not the same issue." + }, + { + "left": "issue:44220", + "right": "issue:44661", + "accept": false, + "reason": "Audio feature extraction issue vs add-new-model-like/tokenizer mapping failure." + }, + { + "left": "issue:44077", + "right": "issue:44625", + "accept": false, + "reason": "PatchTST post_init policy vs Qwen3.5 num_labels propagation; unrelated." + }, + { + "left": "issue:43746", + "right": "issue:46050", + "accept": false, + "reason": "PEFT/local checkpoint loading vs quantized save/load corruption; different problems." + }, + { + "left": "issue:43618", + "right": "issue:43697", + "accept": false, + "reason": "CLIP attentions assignment regression vs RTDetrV2 output differences in v5." + }, + { + "left": "issue:42915", + "right": "issue:43746", + "accept": false, + "reason": "Different models and failure points: Qwen3Moe FP8 config vs GraniteSpeech PEFT loading." + }, + { + "left": "issue:43550", + "right": "issue:45072", + "accept": false, + "reason": "Both touch performance/dtype areas, but the concrete bugs are different models and different failures." + }, + { + "left": "issue:43295", + "right": "issue:45834", + "accept": false, + "reason": "Processor/tokenizer regression vs Kosmos2.5 long-OCR index error; not the same bug." + }, + { + "left": "issue:43550", + "right": "issue:43819", + "accept": false, + "reason": "Unrelated issues: SDPA/torch.compile failure vs DAC latent decoding mismatch." + }, + { + "left": "issue:43540", + "right": "issue:44008", + "accept": false, + "reason": "Both are multimodal, but one is Qwen3Omni video input handling and the other is a Gemma3n variable-name collision." + }, + { + "left": "issue:44186", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer padding/NER crash vs Perceiver image-res interpolation failure; unrelated." + }, + { + "left": "issue:44062", + "right": "issue:44488", + "accept": false, + "reason": "AddedToken keyword conflict vs loading a specific tokenizer model; both are tokenizer-related but not the same defect." + }, + { + "left": "issue:43653", + "right": "issue:43819", + "accept": false, + "reason": "Special-token registration bug vs DAC STE mismatch; no shared code-path." + }, + { + "left": "issue:45183", + "right": "issue:45479", + "accept": false, + "reason": "Validation/error-message bug vs degenerate zero-loss behavior; different bugs and fixes." + }, + { + "left": "issue:43531", + "right": "issue:43643", + "accept": false, + "reason": "Qwen3-MoE sliding-window bug vs AutoConfig trust_remote_code field loss; unrelated." + }, + { + "left": "issue:43335", + "right": "issue:43756", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug vs Smollm3 RoPE-layer dropping; different model/config issues." + }, + { + "left": "issue:43388", + "right": "issue:46032", + "accept": false, + "reason": "Metrics gathering label truncation vs Mamba2 cache correctness; unrelated." + }, + { + "left": "issue:43643", + "right": "issue:44877", + "accept": false, + "reason": "Both are config-loading issues, but they concern different mechanisms and models." + }, + { + "left": "issue:43716", + "right": "issue:44008", + "accept": false, + "reason": "Different multimodal paths and different error types: image dtype mismatch vs audio_mel_mask collision." + }, + { + "left": "issue:44843", + "right": "issue:45081", + "accept": false, + "reason": "Same helper function, but different concrete failures: offline model_info call vs backend_tokenizer AttributeError." + }, + { + "left": "issue:30064", + "right": "issue:36331", + "accept": false, + "reason": "Image segmentation processing vs custom trainer loss signature; unrelated." + }, + { + "left": "issue:43950", + "right": "issue:45987", + "accept": false, + "reason": "Both involve from_pretrained/save behavior, but one is buffer restoration corruption and the other is seq_length return type." + } + ] + }, + "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": 20, + "review_unit_index": 17, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28218", + "issue:29159", + "issue:34634", + "issue:36010", + "issue:39692", + "issue:42175", + "issue:43322", + "issue:43421", + "issue:43452", + "issue:43540", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43720", + "issue:43782", + "issue:43818", + "issue:43825", + "issue:43827", + "issue:43994", + "issue:44186", + "issue:44442", + "issue:44448", + "issue:44479", + "issue:44485", + "issue:44509", + "issue:44561", + "issue:44610", + "issue:44671", + "issue:44805", + "issue:44811", + "issue:44821", + "issue:44857", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45020", + "issue:45072", + "issue:45081", + "issue:45084", + "issue:45381", + "issue:45584", + "issue:45685", + "issue:45834", + "issue:45987", + "issue:46192", + "issue:46291", + "issue:46294", + "issue:46463" + ], + "soft_pairs": [ + "issue:44821|issue:45084", + "issue:44448|issue:44671", + "issue:39692|issue:45834", + "issue:43421|issue:44811", + "issue:43540|issue:45381", + "issue:45584|issue:45834", + "issue:43577|issue:43818", + "issue:43782|issue:44977", + "issue:43606|issue:45072", + "issue:44479|issue:45834", + "issue:44186|issue:44442", + "issue:46294|issue:46463", + "issue:43322|issue:45685", + "issue:43825|issue:43994", + "issue:44485|issue:44509", + "issue:34634|issue:36010", + "issue:44561|issue:45020", + "issue:44991|issue:46291", + "issue:45987|issue:46192", + "issue:43452|issue:44991", + "issue:44805|issue:44936", + "issue:28218|issue:29159", + "issue:44610|issue:45072", + "issue:43720|issue:45072", + "issue:43577|issue:43582", + "issue:43825|issue:43827", + "issue:42175|issue:45081", + "issue:43582|issue:44857" + ], + "prepared_review_unit_hash": "9cf7624fafd615ec615959d2c21fec7c6b9445b14d4ae31dbb8e3ac0b75f3fed", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36187, + "estimated_input_tokens": 9047, + "estimated_eval_tokens": 18350 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly unrelated issues. The only clear duplicate-like pair is the tokenizer added-token spacing bug and the broader slow/fast tokenizer inconsistency report.", + "confidence": 0.94, + "canonical_issue_reason": "issue:29159 is the broader tokenizer inconsistency report and is the best umbrella for the only plausible duplicate pair with 28218.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:29159 best represents the duplicate pair because it states the general slow-vs-fast tokenizer inconsistency, while 28218 is the narrower symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44821", + "right": "issue:45084", + "accept": false, + "reason": "Different bugs: image processor loading from URL vs template compilation error." + }, + { + "left": "issue:44448", + "right": "issue:44671", + "accept": false, + "reason": "Both are v5 regressions, but they affect different models and different failure modes." + }, + { + "left": "issue:39692", + "right": "issue:45834", + "accept": false, + "reason": "Doc-example mistakes vs a runtime OCR index error; not the same bug." + }, + { + "left": "issue:43421", + "right": "issue:44811", + "accept": false, + "reason": "Both involve token/special-token handling, but one is post-processor updates and the other is batch_decode skip_special_tokens behavior." + }, + { + "left": "issue:43540", + "right": "issue:45381", + "accept": false, + "reason": "Different Qwen video-input regressions with different symptoms and model paths." + }, + { + "left": "issue:45584", + "right": "issue:45834", + "accept": false, + "reason": "Whisper empty-transcription failure is unrelated to Kosmos2.5 OCR indexing." + }, + { + "left": "issue:43577", + "right": "issue:43818", + "accept": false, + "reason": "Different models and different loading bugs: dtype propagation vs missing temporal attention/weight sharing." + }, + { + "left": "issue:43782", + "right": "issue:44977", + "accept": false, + "reason": "Qwen3VL weight_only loading error is not the same as Qwen3.5 flash-attention generation failure." + }, + { + "left": "issue:43606", + "right": "issue:45072", + "accept": false, + "reason": "Different root causes: CPU offload device mismatch vs bfloat16 dtype mismatch." + }, + { + "left": "issue:44479", + "right": "issue:45834", + "accept": false, + "reason": "Unrelated: Qwen video regression vs Kosmos2.5 long OCR input error." + }, + { + "left": "issue:44186", + "right": "issue:44442", + "accept": false, + "reason": "Tokenizer crash on NER/padding is unrelated to AutoTokenizer loading a FastSpeech2 tokenizer." + }, + { + "left": "issue:46294", + "right": "issue:46463", + "accept": false, + "reason": "Qwen video feature/token mismatch is unrelated to T5 decoder-input validation." + }, + { + "left": "issue:43322", + "right": "issue:45685", + "accept": false, + "reason": "Llava Next load segfault and MPS histogram issue are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43994", + "accept": false, + "reason": "Pipeline error-message text and SigLIP2 output quality are different issues." + }, + { + "left": "issue:44485", + "right": "issue:44509", + "accept": false, + "reason": "GLM-5 RoPE implementation is unrelated to pipeline-task documentation drift." + }, + { + "left": "issue:34634", + "right": "issue:36010", + "accept": false, + "reason": "BarkProcessor voice_preset and GenerationMixin import failure are unrelated." + }, + { + "left": "issue:44561", + "right": "issue:45020", + "accept": false, + "reason": "Related theme, but one is a specific helper removal and the other is a broad remote_code breakage report." + }, + { + "left": "issue:44991", + "right": "issue:46291", + "accept": false, + "reason": "Tokenizer loading failure and kernels import/version requirement failure are unrelated." + }, + { + "left": "issue:45987", + "right": "issue:46192", + "accept": false, + "reason": "StaticCache seq-length contract bug and return_tensors kwargs handling are unrelated." + }, + { + "left": "issue:43452", + "right": "issue:44991", + "accept": false, + "reason": "Both are loading-related, but gguf_file breakage and a specific tokenizer load failure are different bugs." + }, + { + "left": "issue:44805", + "right": "issue:44936", + "accept": false, + "reason": "Mask-shape IndexError and trainer.evaluate() after train are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:29159", + "accept": true, + "reason": "Added-token extra spacing matches the broader slow-vs-fast tokenizer inconsistency report." + }, + { + "left": "issue:44610", + "right": "issue:45072", + "accept": false, + "reason": "Processor input-size mismatch and dtype mismatch are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:45072", + "accept": false, + "reason": "Packed-weight loading failure is not the same as a bfloat16 dtype mismatch." + }, + { + "left": "issue:43577", + "right": "issue:43582", + "accept": false, + "reason": "Blip2 dtype propagation and AppleSilicon allocator warmup TypeError are unrelated." + }, + { + "left": "issue:43825", + "right": "issue:43827", + "accept": false, + "reason": "Both involve v5 pipeline removal, but one is an error message and the other is docs drift." + }, + { + "left": "issue:42175", + "right": "issue:45081", + "accept": false, + "reason": "TF backend packaging and Mistral regex patch crash are unrelated." + }, + { + "left": "issue:43582", + "right": "issue:44857", + "accept": false, + "reason": "AppleSilicon warmup TypeError and CUDA AMP loss crash are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only plausible duplicate-like pair is issue:28218 with issue:29159, and the rest of the soft-edge rejections are generally supported by the issue titles and descriptions. No PR-pair mergeability concerns apply here." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 18, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:34634", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:42175", + "issue:43262", + "issue:43329", + "issue:43404", + "issue:43526", + "issue:43550", + "issue:43638", + "issue:43644", + "issue:43645", + "issue:43650", + "issue:43653", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43746", + "issue:43749", + "issue:43782", + "issue:43819", + "issue:43854", + "issue:43906", + "issue:43927", + "issue:44060", + "issue:44368", + "issue:44589", + "issue:44610", + "issue:44671", + "issue:44811", + "issue:44912", + "issue:44936", + "issue:44977", + "issue:45003", + "issue:45020", + "issue:45042", + "issue:45081", + "issue:45084", + "issue:45406", + "issue:45538", + "issue:45600", + "issue:45701", + "issue:46209", + "issue:46285", + "issue:46291", + "issue:46421", + "issue:46424" + ], + "soft_pairs": [ + "issue:43404|issue:43782", + "issue:43262|issue:43819", + "issue:44671|issue:45701", + "issue:45003|issue:46285", + "issue:43526|issue:43697", + "issue:43653|issue:44912", + "issue:45081|issue:45406", + "issue:44060|issue:44368", + "issue:43906|issue:46209", + "issue:43742|issue:45020", + "issue:45084|issue:45538", + "issue:43329|issue:45600", + "issue:44589|issue:45042", + "issue:43644|issue:43749", + "issue:43550|issue:44912", + "issue:43638|issue:44936", + "issue:43746|issue:44811", + "issue:43819|issue:43927", + "issue:34634|issue:36331", + "issue:43650|issue:43906", + "issue:36010|issue:43645", + "issue:38175|issue:44610", + "issue:42175|issue:46291", + "issue:43720|issue:44610", + "issue:46421|issue:46424", + "issue:43854|issue:44977", + "issue:36010|issue:36331" + ], + "prepared_review_unit_hash": "f354d86848190bd3ffc31745044d19453ae7fcca485c6193c00063dcdec2f940", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35970, + "estimated_input_tokens": 8993, + "estimated_eval_tokens": 18242 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This set is mostly a grab bag of unrelated Transformers issues. None of the soft-similarity pairs look like true duplicates or the same underlying fix, so I reject all soft edges.", + "confidence": 0.95, + "canonical_issue_reason": "issue:45020 is the broadest concrete regression report in the set, but there is no strong single canonical because the items are largely unrelated.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 is the closest thing to an umbrella anchor among these issues, but only weakly so given the heterogeneity of the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43404", + "right": "issue:43782", + "accept": false, + "reason": "Different models and different failures: tied lm_head weights vs a weight_only loading error." + }, + { + "left": "issue:43262", + "right": "issue:43819", + "accept": false, + "reason": "Audio chat-template sampling-rate bug is unrelated to DAC latent reconstruction mismatch." + }, + { + "left": "issue:44671", + "right": "issue:45701", + "accept": false, + "reason": "CamemBERT masked-LM regression is not the same as a generic tokenization-change report." + }, + { + "left": "issue:45003", + "right": "issue:46285", + "accept": false, + "reason": "Unsafe sys.modules access is unrelated to CLIP structural inconsistency." + }, + { + "left": "issue:43526", + "right": "issue:43697", + "accept": false, + "reason": "BEiT label reduction bug and RTDetr output drift are different code paths and symptoms." + }, + { + "left": "issue:43653", + "right": "issue:44912", + "accept": false, + "reason": "BigBirdTokenizer decode/special-token issue is unrelated to MXFP4 quantization loading fallback." + }, + { + "left": "issue:45081", + "right": "issue:45406", + "accept": false, + "reason": "Both are load-time processor/tokenizer failures, but different models and different missing attributes/code paths." + }, + { + "left": "issue:44060", + "right": "issue:44368", + "accept": false, + "reason": "Both mention tied-weight warnings, but they concern different Qwen variants and different warning behavior." + }, + { + "left": "issue:43906", + "right": "issue:46209", + "accept": false, + "reason": "A reproduction of an older issue is not the same as an FP8 inference failure on B200." + }, + { + "left": "issue:43742", + "right": "issue:45020", + "accept": false, + "reason": "MobileLLM key error and remote_code regressions are different loading problems." + }, + { + "left": "issue:45084", + "right": "issue:45538", + "accept": false, + "reason": "Template-node compilation failure is unrelated to CLIPTokenizer's model_max_length bug." + }, + { + "left": "issue:43329", + "right": "issue:45600", + "accept": false, + "reason": "Undefined multimodal-token variables and removed auto-mapping configs are separate regressions." + }, + { + "left": "issue:44589", + "right": "issue:45042", + "accept": false, + "reason": "Float8 storage lookup error is not the same as PIL backend requiring torchvision." + }, + { + "left": "issue:43644", + "right": "issue:43749", + "accept": false, + "reason": "Junk non-persistent buffers and broken FSDP CPU RAM loading are different initialization/loading defects." + }, + { + "left": "issue:43550", + "right": "issue:44912", + "accept": false, + "reason": "Bamba torch.compile+SDPA failure is unrelated to MXFP4 quantization fallback." + }, + { + "left": "issue:43638", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO3 index error during training is not the same as evaluate-after-train failure." + }, + { + "left": "issue:43746", + "right": "issue:44811", + "accept": false, + "reason": "PEFT local-checkpoint loading and Whisper skip_special_tokens decoding are unrelated." + }, + { + "left": "issue:43819", + "right": "issue:43927", + "accept": false, + "reason": "DAC forward/latent mismatch is unrelated to DiaConfig losing custom token IDs on save/load." + }, + { + "left": "issue:34634", + "right": "issue:36331", + "accept": false, + "reason": "BarkProcessor voice_preset and CustomTrainer.compute_loss signature errors are unrelated." + }, + { + "left": "issue:43650", + "right": "issue:43906", + "accept": false, + "reason": "An empty placeholder issue is not a duplicate of an isolated reproduction report." + }, + { + "left": "issue:36010", + "right": "issue:43645", + "accept": false, + "reason": "GenerationMixin import breakage is unrelated to notebook-defined custom model initialization." + }, + { + "left": "issue:38175", + "right": "issue:44610", + "accept": false, + "reason": "Siglip2 zero-probabilities and OmDet processor size mismatch are different model-specific bugs." + }, + { + "left": "issue:42175", + "right": "issue:46291", + "accept": false, + "reason": "Missing tensorflow in an install extra and kernels import requirements are separate dependency issues." + }, + { + "left": "issue:43720", + "right": "issue:44610", + "accept": false, + "reason": "Packed-weight loading failure for BitNet is unrelated to OmDet-Turbo image-size mismatch." + }, + { + "left": "issue:46421", + "right": "issue:46424", + "accept": false, + "reason": "Both are cache-related, but one is lazy-init recompilation and the other is max_cache_length handling." + }, + { + "left": "issue:43854", + "right": "issue:44977", + "accept": false, + "reason": "GLM-4.7 unit-test loading and Qwen3.5 flash-attention generation are different model-specific regressions." + }, + { + "left": "issue:36010", + "right": "issue:36331", + "accept": false, + "reason": "ImportError for GenerationMixin and CustomTrainer.compute_loss keyword mismatch are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 19, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:29159", + "issue:34567", + "issue:39692", + "issue:40990", + "issue:42490", + "issue:42915", + "issue:43257", + "issue:43262", + "issue:43329", + "issue:43404", + "issue:43425", + "issue:43450", + "issue:43452", + "issue:43454", + "issue:43525", + "issue:43526", + "issue:43531", + "issue:43577", + "issue:43582", + "issue:43606", + "issue:43643", + "issue:43697", + "issue:43742", + "issue:43749", + "issue:43927", + "issue:43931", + "issue:44291", + "issue:44387", + "issue:44479", + "issue:44554", + "issue:44568", + "issue:44617", + "issue:44811", + "issue:44821", + "issue:44849", + "issue:44912", + "issue:45042", + "issue:45127", + "issue:45356", + "issue:45357", + "issue:45372", + "issue:45459", + "issue:45701", + "issue:45759", + "issue:45803", + "issue:45859", + "issue:46291", + "issue:46326" + ], + "soft_pairs": [ + "issue:39692|issue:44912", + "issue:44387|issue:44617", + "issue:43531|issue:45042", + "issue:45459|issue:45803", + "issue:43526|issue:44554", + "issue:43749|issue:44291", + "issue:43262|issue:44811", + "issue:43257|issue:43329", + "issue:43643|issue:45759", + "issue:43454|issue:43577", + "issue:45372|issue:46326", + "issue:42490|issue:43257", + "issue:43329|issue:43450", + "issue:43606|issue:45859", + "issue:34567|issue:40990", + "issue:43329|issue:44479", + "issue:44568|issue:45356", + "issue:42915|issue:43742", + "issue:29159|issue:34567", + "issue:43525|issue:46291", + "issue:43404|issue:45127", + "issue:44291|issue:45042", + "issue:43257|issue:43931", + "issue:43927|issue:44912", + "issue:42915|issue:44821", + "issue:43697|issue:45357", + "issue:43452|issue:43582", + "issue:44849|issue:45701", + "issue:43425|issue:46291" + ], + "prepared_review_unit_hash": "b27b6bfaca451adaac5e12953306dd54b4c815fb0abfd123b825f47326d67df6", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36724, + "estimated_input_tokens": 9181, + "estimated_eval_tokens": 18618 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All candidate links are false positives: they span unrelated tokenizer, model-loading, quantization, processor, and CPU-offload bugs. No pair looks like the same underlying issue/change, so there is no duplicate cluster to collapse.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:39692", + "right": "issue:44912", + "accept": false, + "reason": "Docs/example breakage in SigLIP2 is unrelated to gpt-oss MXFP4 quantization fallback." + }, + { + "left": "issue:44387", + "right": "issue:44617", + "accept": false, + "reason": "Both mention OOM, but one is int4 CUDA reserved-memory growth and the other is Sam3Video-specific CPU/GPU usage." + }, + { + "left": "issue:43531", + "right": "issue:45042", + "accept": false, + "reason": "Qwen3-MoE sliding_window logic and PIL backend torchvision dependency are different bugs." + }, + { + "left": "issue:45459", + "right": "issue:45803", + "accept": false, + "reason": "Different exception-handling paths: protobuf decode masking vs bare except in Fuyu tensor conversion." + }, + { + "left": "issue:43526", + "right": "issue:44554", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and MPS attention correctness are unrelated code paths." + }, + { + "left": "issue:43749", + "right": "issue:44291", + "accept": false, + "reason": "Both concern loading, but FSDP CPU RAM-efficient loading and init_empty_weights argument handling are distinct failures." + }, + { + "left": "issue:43262", + "right": "issue:44811", + "accept": false, + "reason": "Audio chat-template sampling-rate defaulting is unrelated to Whisper batch_decode skipping special tokens." + }, + { + "left": "issue:43257", + "right": "issue:43329", + "accept": false, + "reason": "Qwen3 MoE weight conversion under accelerate/deepspeed is not the same as undefined video token-count helpers." + }, + { + "left": "issue:43643", + "right": "issue:45759", + "accept": false, + "reason": "AutoConfig remote-code field loss and composite Qwen text_config unwrapping are different API bugs." + }, + { + "left": "issue:43454", + "right": "issue:43577", + "accept": false, + "reason": "Weight tying in AyaVision/Mistral3 is unrelated to Blip2 dtype propagation." + }, + { + "left": "issue:45372", + "right": "issue:46326", + "accept": false, + "reason": "One is an import failure from mistral_common, the other is continue_final_message behavior in a tokenizer." + }, + { + "left": "issue:42490", + "right": "issue:43257", + "accept": false, + "reason": "Local-path trailing-slash warning is unrelated to Qwen3 MoE loading under accelerate/deepspeed." + }, + { + "left": "issue:43329", + "right": "issue:43450", + "accept": false, + "reason": "Undefined video patch helpers and batched video shape errors are separate video-processing bugs." + }, + { + "left": "issue:43606", + "right": "issue:45859", + "accept": false, + "reason": "Both involve CPU offload, but they affect different models and code paths." + }, + { + "left": "issue:34567", + "right": "issue:40990", + "accept": false, + "reason": "Trainer token-count tracking does not match a perplexity regression on gpt-oss." + }, + { + "left": "issue:43329", + "right": "issue:44479", + "accept": false, + "reason": "A missing helper in multimodal token counting is not the same as the v5.3.0 video input regression." + }, + { + "left": "issue:44568", + "right": "issue:45356", + "accept": false, + "reason": "Model-specific tokenizer special-token handling and Kimi codec handling are different regressions." + }, + { + "left": "issue:42915", + "right": "issue:43742", + "accept": false, + "reason": "FineGrainedFP8Config failure on Qwen3 MoE is unrelated to MobileLLM key errors." + }, + { + "left": "issue:29159", + "right": "issue:34567", + "accept": false, + "reason": "Tokenizer parity issues and TrainerState token counters are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:46291", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is unrelated to kernels package import/version requirements." + }, + { + "left": "issue:43404", + "right": "issue:45127", + "accept": false, + "reason": "Missing lm_head tying and LoRA-extended-vocabulary collapse are different weight-sync problems." + }, + { + "left": "issue:44291", + "right": "issue:45042", + "accept": false, + "reason": "init_empty_weights argument compatibility has nothing to do with PIL backend torchvision requirements." + }, + { + "left": "issue:43257", + "right": "issue:43931", + "accept": false, + "reason": "Qwen3 MoE loading under accelerate/deepspeed is not the same as Qwen3-VL weight-shape mismatch." + }, + { + "left": "issue:43927", + "right": "issue:44912", + "accept": false, + "reason": "DiaConfig save/load token IDs loss and git-oss quantization fallback are unrelated." + }, + { + "left": "issue:42915", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3 MoE FP8 loading and AutoImageProcessor URL loading are unrelated issues." + }, + { + "left": "issue:43697", + "right": "issue:45357", + "accept": false, + "reason": "RTDetrV2 output drift and Qwen3.5 save_pretrained visual-key corruption are different bugs." + }, + { + "left": "issue:43452", + "right": "issue:43582", + "accept": false, + "reason": "gguf_file loading failures and Apple Silicon warmup TypeError are unrelated." + }, + { + "left": "issue:44849", + "right": "issue:45701", + "accept": false, + "reason": "Qwen3.5 hidden-state output bug and version-dependent tokenization changes are different regressions." + }, + { + "left": "issue:43425", + "right": "issue:46291", + "accept": false, + "reason": "Torch 2.10 compatibility and kernels v0.15 import requirements are separate dependency issues." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 20, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:24643", + "issue:27957", + "issue:28282", + "issue:29159", + "issue:33357", + "issue:34567", + "issue:36010", + "issue:36331", + "issue:38175", + "issue:41720", + "issue:42175", + "issue:42915", + "issue:43257", + "issue:43317", + "issue:43404", + "issue:43493", + "issue:43540", + "issue:43577", + "issue:43646", + "issue:43696", + "issue:43742", + "issue:43749", + "issue:43906", + "issue:43994", + "issue:44062", + "issue:44291", + "issue:44479", + "issue:44560", + "issue:44568", + "issue:44737", + "issue:44792", + "issue:44898", + "issue:44964", + "issue:44987", + "issue:44991", + "issue:45072", + "issue:45081", + "issue:45137", + "issue:45381", + "issue:45405", + "issue:45488", + "issue:45538", + "issue:45588", + "issue:45907", + "issue:45987", + "issue:46463", + "issue:46468", + "issue:46491" + ], + "soft_pairs": [ + "issue:42915|issue:45538", + "issue:44987|issue:45538", + "issue:41720|issue:43540", + "issue:43317|issue:45538", + "issue:33357|issue:34567", + "issue:45137|issue:45405", + "issue:43540|issue:44479", + "issue:43540|issue:44560", + "issue:44568|issue:45488", + "issue:44479|issue:45381", + "issue:43742|issue:44291", + "issue:44987|issue:45907", + "issue:24643|issue:29159", + "issue:38175|issue:44898", + "issue:44737|issue:46463", + "issue:44062|issue:44991", + "issue:43257|issue:43696", + "issue:43906|issue:46468", + "issue:34567|issue:36331", + "issue:43646|issue:43749", + "issue:43577|issue:45072", + "issue:44568|issue:46491", + "issue:44737|issue:45588", + "issue:44062|issue:45081", + "issue:43493|issue:43994", + "issue:27957|issue:28282", + "issue:43404|issue:45987", + "issue:44792|issue:44964", + "issue:36010|issue:42175" + ], + "prepared_review_unit_hash": "e241ea402059b5927ad5fa6f3990a9da798f97f0e0d30cb6fd821c49157511fb", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36510, + "estimated_input_tokens": 9128, + "estimated_eval_tokens": 18512 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous cluster of unrelated Transformers bugs, with only a few plausible duplicate groups: Qwen 5.3.0 video-input regressions and SigLIP2 output discrepancies. No PRs are present.", + "confidence": 0.81, + "canonical_issue_reason": "issue:44479 is the broadest representative of the strongest overlap in this set: the Qwen 5.3.0 video-input regression family. It covers the shared multimodal/video code path that also appears in the more specific Qwen reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44479 is the best single issue anchor because it is the most umbrella-like, recent, and central to the only clear duplicate cluster (Qwen video-input regressions).", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42915", + "right": "issue:45538", + "accept": false, + "reason": "Different subsystems and failure modes: Qwen3MoE FP8 loading vs CLIPTokenizer max length." + }, + { + "left": "issue:44987", + "right": "issue:45538", + "accept": false, + "reason": "Unrelated: a model-loading failure for physical-intelligence/fast versus a CLIP tokenizer constant issue." + }, + { + "left": "issue:41720", + "right": "issue:43540", + "accept": false, + "reason": "Both involve Qwen models, but one is auto device mapping/cuda assert and the other is video-input processing; not the same bug." + }, + { + "left": "issue:43317", + "right": "issue:45538", + "accept": false, + "reason": "Different failures: dequantized model offload loading versus tokenizer max-length behavior." + }, + { + "left": "issue:33357", + "right": "issue:34567", + "accept": false, + "reason": "MacOS bus error on CLIP versus TrainerState token counter not updating; unrelated." + }, + { + "left": "issue:45137", + "right": "issue:45405", + "accept": false, + "reason": "DeepSpeed ZeRO3 deque error and a PEFT version bump are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:44479", + "accept": true, + "reason": "Same Qwen video-input regression family; 44479 is the umbrella report and 43540 is a specific model symptom." + }, + { + "left": "issue:43540", + "right": "issue:44560", + "accept": true, + "reason": "Both are Qwen multimodal video-input failures in the same regression family, likely the same processing bug." + }, + { + "left": "issue:44568", + "right": "issue:45488", + "accept": false, + "reason": "Different tokenizer regressions affecting different models and code paths." + }, + { + "left": "issue:44479", + "right": "issue:45381", + "accept": true, + "reason": "Same Qwen VL video regression area; 45381 looks like a specific manifestation of the broader 5.3.0 issue." + }, + { + "left": "issue:43742", + "right": "issue:44291", + "accept": false, + "reason": "Both are model loading errors, but one is a key error for a specific model and the other is an init_empty_weights API break; not the same bug." + }, + { + "left": "issue:44987", + "right": "issue:45907", + "accept": false, + "reason": "Loading a model from the Hub and forwarding list_repo_files revision are unrelated." + }, + { + "left": "issue:24643", + "right": "issue:29159", + "accept": false, + "reason": "DeepSpeed training weight-shape error and tokenizer inconsistency are unrelated." + }, + { + "left": "issue:38175", + "right": "issue:44898", + "accept": false, + "reason": "Both are vision-model bugs, but they concern different architectures and symptoms." + }, + { + "left": "issue:44737", + "right": "issue:46463", + "accept": false, + "reason": "XLNet CPU placement bug and T5 decoder error-message bug are unrelated." + }, + { + "left": "issue:44062", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading issues, but the titles point to different root causes and model families." + }, + { + "left": "issue:43257", + "right": "issue:43696", + "accept": false, + "reason": "Qwen3 MoE weight-conversion bug versus GPT-oss OOM; unrelated." + }, + { + "left": "issue:43906", + "right": "issue:46468", + "accept": false, + "reason": "Different models and different failures; no clear shared code-path bug." + }, + { + "left": "issue:34567", + "right": "issue:36331", + "accept": false, + "reason": "TrainerState token accounting and CustomTrainer compute_loss signature mismatch are distinct trainer issues." + }, + { + "left": "issue:43646", + "right": "issue:43749", + "accept": false, + "reason": "Custom model initialization break and FSDP CPU RAM efficient loading are separate regressions." + }, + { + "left": "issue:43577", + "right": "issue:45072", + "accept": false, + "reason": "Both mention dtype mismatches, but they affect different model families and likely different code paths." + }, + { + "left": "issue:44568", + "right": "issue:46491", + "accept": false, + "reason": "Different tokenizer regressions for different models." + }, + { + "left": "issue:44737", + "right": "issue:45588", + "accept": false, + "reason": "XLNet positional encoding CPU placement and flash-attention crash on sink-less models are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43994", + "accept": true, + "reason": "Both are SigLIP2 correctness regressions: one frames it as a HF-vs-JAX discrepancy, the other as nonsensical outputs on the same model family." + }, + { + "left": "issue:27957", + "right": "issue:28282", + "accept": false, + "reason": "Flash Attention 2 on XLM-Roberta versus missing PyTorch import are unrelated." + }, + { + "left": "issue:43404", + "right": "issue:45987", + "accept": false, + "reason": "Tied lm_head weights in Mistral3 and StaticCache length return type are unrelated." + }, + { + "left": "issue:44792", + "right": "issue:44964", + "accept": false, + "reason": "Different multimodal models and different failures; not the same bug." + }, + { + "left": "issue:36010", + "right": "issue:42175", + "accept": false, + "reason": "GenerationMixin import error and TensorFlow backend packaging are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well grounded in the packet and stays conservative. The Qwen video-input issues are the clearest overlap, and the SigLIP2 pair is a plausible second cluster. No PRs are present, and the soft-edge verdicts do not overclaim beyond the metadata." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 21, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28282", + "issue:29127", + "issue:34689", + "issue:41762", + "issue:42175", + "issue:42898", + "issue:43257", + "issue:43452", + "issue:43526", + "issue:43531", + "issue:43582", + "issue:43643", + "issue:43697", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43756", + "issue:43824", + "issue:43906", + "issue:44291", + "issue:44488", + "issue:44589", + "issue:44625", + "issue:44756", + "issue:44821", + "issue:44877", + "issue:44898", + "issue:44913", + "issue:44936", + "issue:44977", + "issue:45042", + "issue:45081", + "issue:45200", + "issue:45216", + "issue:45325", + "issue:45357", + "issue:45375", + "issue:45381", + "issue:45405", + "issue:45479", + "issue:45600", + "issue:45823", + "issue:45920", + "issue:46121", + "issue:46399", + "issue:46463", + "issue:46491" + ], + "soft_pairs": [ + "issue:44291|issue:44821", + "issue:28282|issue:34689", + "issue:42175|issue:43824", + "issue:43749|issue:44756", + "issue:45823|issue:46463", + "issue:45042|issue:45081", + "issue:43742|issue:44589", + "issue:43257|issue:43756", + "issue:43452|issue:44488", + "issue:43906|issue:45381", + "issue:45405|issue:45600", + "issue:45357|issue:45920", + "issue:28282|issue:29127", + "issue:43531|issue:44821", + "issue:43756|issue:46399", + "issue:44913|issue:46121", + "issue:43526|issue:45479", + "issue:43643|issue:45375", + "issue:43697|issue:45216", + "issue:43720|issue:44898", + "issue:45200|issue:46463", + "issue:44625|issue:44877", + "issue:43756|issue:45325", + "issue:42898|issue:46491", + "issue:43582|issue:44977", + "issue:41762|issue:44936" + ], + "prepared_review_unit_hash": "0b09a0db05cc6b9f483a7c42be4182a76e87abab3620bf948276207888f69ab3", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 26, + "serialized_chars": 35265, + "estimated_input_tokens": 8817, + "estimated_eval_tokens": 17890 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of unrelated model-loading, config, tokenizer, and regression bugs; I don\u2019t see any soft pair that looks like the same underlying defect, so all soft edges should be rejected. If a single anchor is needed, issue 45920 is the broadest, most general regression report in the set, but the cluster is not a true duplicate group.", + "confidence": 0.91, + "canonical_issue_reason": "Issue 45920 is the broadest and most central report here: it describes a multi-model tokenizer regression with wide blast radius, whereas the others are narrower, model-specific bugs or unrelated failures.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 45920 is the best representative issue if one must be chosen, because its scope is broad and it reads like a family-level regression rather than a single isolated model bug.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44291", + "right": "issue:44821", + "accept": false, + "reason": "Different code paths: init_empty_weights/_is_hf_initialized argument handling vs loading AutoImageProcessor from a URL." + }, + { + "left": "issue:28282", + "right": "issue:34689", + "accept": false, + "reason": "A missing PyTorch import error is not the same as a Llama 3.2 Vision model-loading regression." + }, + { + "left": "issue:42175", + "right": "issue:43824", + "accept": false, + "reason": "Package/backend dependency mismatch vs missing Qwen2_5_VL import; unrelated failures." + }, + { + "left": "issue:43749", + "right": "issue:44756", + "accept": false, + "reason": "Both are loading/memory-related, but one is FSDP CPU-efficient loading and the other is mmap/OOM on Strix Halo; not the same bug." + }, + { + "left": "issue:45823", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 device_map forward mismatch and T5 decoder input validation are different subsystems and symptoms." + }, + { + "left": "issue:45042", + "right": "issue:45081", + "accept": false, + "reason": "PIL backend dependency regression vs Mistral tokenizer regex crash; no shared underlying defect." + }, + { + "left": "issue:43742", + "right": "issue:44589", + "accept": false, + "reason": "MobileLLM key-error loading failure vs missing Float8 storage type during deserialization; different causes." + }, + { + "left": "issue:43257", + "right": "issue:43756", + "accept": false, + "reason": "Qwen3 MoE weight conversion/loading bug vs Smollm3 RoPE-layer selection bug; not the same issue." + }, + { + "left": "issue:43452", + "right": "issue:44488", + "accept": false, + "reason": "gguf_file breaks tokenizer/model loading, but cjvt/sleng-bert is a separate model-loading failure." + }, + { + "left": "issue:43906", + "right": "issue:45381", + "accept": false, + "reason": "One is a reproduction of an older issue; the other is a Qwen2.5-VL video position_ids bug." + }, + { + "left": "issue:45405", + "right": "issue:45600", + "accept": false, + "reason": "A version-bump/release-management issue vs stale auto_mappings references; unrelated." + }, + { + "left": "issue:45357", + "right": "issue:45920", + "accept": false, + "reason": "Qwen3.5 save_pretrained visual-key regression is different from a broad AutoTokenizer token-ID regression." + }, + { + "left": "issue:28282", + "right": "issue:29127", + "accept": false, + "reason": "Missing PyTorch import handling is unrelated to LayoutLMv3 box-information error messaging." + }, + { + "left": "issue:43531", + "right": "issue:44821", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and AutoImageProcessor URL loading are unrelated." + }, + { + "left": "issue:43756", + "right": "issue:46399", + "accept": false, + "reason": "RoPE-layer dropping in Smollm3 is not the same as a missing decorator causing kernel registration failure." + }, + { + "left": "issue:44913", + "right": "issue:46121", + "accept": false, + "reason": "Config reload losing rotary_pct is separate from JSON-loaded list handling in convert_rope_params_to_dict." + }, + { + "left": "issue:43526", + "right": "issue:45479", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug and zero-loss single-label classification bug are different behaviors." + }, + { + "left": "issue:43643", + "right": "issue:45375", + "accept": false, + "reason": "trust_remote_code missing fields and strict-dropping deepstack_visual_indexes are different config-serialization issues." + }, + { + "left": "issue:43697", + "right": "issue:45216", + "accept": false, + "reason": "RTDetrV2 output drift in v5 is not the same as Qwen3.5 save_pretrained saving incorrect checkpoint keys." + }, + { + "left": "issue:43720", + "right": "issue:44898", + "accept": false, + "reason": "Packed-weight unpacking during accelerate loading is unrelated to Perceiver non-default resolution failure." + }, + { + "left": "issue:45200", + "right": "issue:46463", + "accept": false, + "reason": "Gemma4 mm_token_type_ids defaulting and T5 decoder validation errors are different bugs." + }, + { + "left": "issue:44625", + "right": "issue:44877", + "accept": false, + "reason": "num_labels propagation in Qwen3.5 is separate from granite_speech being blocked by strict config loading." + }, + { + "left": "issue:43756", + "right": "issue:45325", + "accept": false, + "reason": "Different Qwen/Smollm visual-position logic; one is RoPE-layer omission, the other is still-image temporal position_ids scaling." + }, + { + "left": "issue:42898", + "right": "issue:46491", + "accept": false, + "reason": "Tokenization-space cleanup behavior in v5 and CodeLlama round-trip spacing regression are not the same defect." + }, + { + "left": "issue:43582", + "right": "issue:44977", + "accept": false, + "reason": "AppleSilicon warmup TypeError vs Qwen3.5 flash-attention generation bug; different execution paths." + }, + { + "left": "issue:41762", + "right": "issue:44936", + "accept": false, + "reason": "DeepSpeed ZeRO-3 loading IndexError is unrelated to trainer.evaluate() failing after train() due to trainer state." + } + ] + }, + "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": 20, + "review_unit_index": 22, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:24643", + "issue:28282", + "issue:29127", + "issue:29159", + "issue:33357", + "issue:35141", + "issue:38175", + "issue:43295", + "issue:43452", + "issue:43531", + "issue:43550", + "issue:43577", + "issue:43643", + "issue:43653", + "issue:43697", + "issue:43742", + "issue:43756", + "issue:43819", + "issue:43927", + "issue:44361", + "issue:44442", + "issue:44464", + "issue:44466", + "issue:44484", + "issue:44493", + "issue:44514", + "issue:44561", + "issue:44568", + "issue:44704", + "issue:44743", + "issue:44811", + "issue:44843", + "issue:44936", + "issue:44977", + "issue:44991", + "issue:45005", + "issue:45290", + "issue:45356", + "issue:45372", + "issue:45381", + "issue:45491", + "issue:45542", + "issue:45823", + "issue:46002", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46402" + ], + "soft_pairs": [ + "issue:44466|issue:45005", + "issue:24643|issue:29127", + "issue:43452|issue:44843", + "issue:44514|issue:45381", + "issue:43819|issue:44811", + "issue:45372|issue:46170", + "issue:44361|issue:44442", + "issue:38175|issue:43295", + "issue:44936|issue:45823", + "issue:43653|issue:44361", + "issue:44484|issue:44704", + "issue:43452|issue:43577", + "issue:45542|issue:46178", + "issue:43295|issue:45356", + "issue:43577|issue:43697", + "issue:44493|issue:46050", + "issue:43653|issue:44568", + "issue:44843|issue:44991", + "issue:43756|issue:46402", + "issue:29159|issue:33357", + "issue:43531|issue:43742", + "issue:43653|issue:43927", + "issue:44464|issue:44743", + "issue:44514|issue:45290", + "issue:43550|issue:43927", + "issue:43550|issue:44361", + "issue:44464|issue:44977", + "issue:28282|issue:35141", + "issue:43643|issue:44561", + "issue:45491|issue:46002" + ], + "prepared_review_unit_hash": "9421f9274c5c9dbdacfd10bd437ecb8a5a86c9783acfb6b6382608170915cf82", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 37017, + "estimated_input_tokens": 9255, + "estimated_eval_tokens": 18766 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "A heterogeneous cluster of mostly unrelated closed issues around tokenizers, model loading, generation, backend imports, and regression edge cases. No soft pair looks like the same underlying bug/change, so all candidate duplicates are rejected. If forced to pick a representative issue, issue:43452 is the broadest loading-path regression in the set.", + "confidence": 0.9, + "canonical_issue_reason": "issue:43452 is the closest cluster anchor because it covers a broad from_pretrained loading-path regression affecting both tokenizer and causal LM loading.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43452 is the most representative of the cluster\u2019s shared theme of model-loading/tokenizer regressions while remaining a concrete bug report.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44466", + "right": "issue:45005", + "accept": false, + "reason": "Both mention tied weights, but one is a device-dependent lm_head serialization inconsistency and the other is a translation-model tied-weights problem; different failures and likely different fixes." + }, + { + "left": "issue:24643", + "right": "issue:29127", + "accept": false, + "reason": "DeepSpeed training weight-shape runtime error vs LayoutLMv3 box validation/error-message clarity; unrelated code paths." + }, + { + "left": "issue:43452", + "right": "issue:44843", + "accept": false, + "reason": "Both hit AutoTokenizer/from_pretrained, but one is broken by gguf_file handling and the other by an unconditional model_info() call in a regex patch; different bugs." + }, + { + "left": "issue:44514", + "right": "issue:45381", + "accept": false, + "reason": "Both involve Qwen2.5-VL, but one is batched apply_chat_template padding behavior and the other is incorrect video vision_position_ids; not the same bug." + }, + { + "left": "issue:43819", + "right": "issue:44811", + "accept": false, + "reason": "DAC.from_latents vs Whisper batch_decode(skip_special_tokens); no shared underlying defect." + }, + { + "left": "issue:45372", + "right": "issue:46170", + "accept": false, + "reason": "Both are import/dependency failures, but they concern different packages, platforms, and missing symbols." + }, + { + "left": "issue:44361", + "right": "issue:44442", + "accept": false, + "reason": "Different tokenizers and different failures: MLukeTokenizer AttributeError vs AutoTokenizer failing to load FastSpeech2ConformerTokenizer." + }, + { + "left": "issue:38175", + "right": "issue:43295", + "accept": false, + "reason": "Siglip2 zero-probability output is unrelated to the processor.tokenizer regression in v4.57.5." + }, + { + "left": "issue:44936", + "right": "issue:45823", + "accept": false, + "reason": "trainer.evaluate() after train is a training-loop issue, while Gemma4 PLE device mismatch is a model/device placement bug." + }, + { + "left": "issue:43653", + "right": "issue:44361", + "accept": false, + "reason": "Both concern tokenizer behavior, but one is BigBird special-token decode registration and the other is an MLukeTokenizer AttributeError." + }, + { + "left": "issue:44484", + "right": "issue:44704", + "accept": false, + "reason": "save_pretrained shard-size question is unrelated to AutoProcessor not forwarding kwargs to cached_file." + }, + { + "left": "issue:43452", + "right": "issue:43577", + "accept": false, + "reason": "gguf_file loading breakage is unrelated to Blip2 dtype propagation during loading." + }, + { + "left": "issue:45542", + "right": "issue:46178", + "accept": false, + "reason": "Both touch backend selection, but one is undefined tf backend behavior and the other is an invalid tensorflow_text backend mapping." + }, + { + "left": "issue:43295", + "right": "issue:45356", + "accept": false, + "reason": "Processor/tokenizer regression and Kimi tokenizer codec/warning regression are different code paths and symptoms." + }, + { + "left": "issue:43577", + "right": "issue:43697", + "accept": false, + "reason": "Blip2 dtype loading and RTDetrV2 output differences are unrelated model bugs." + }, + { + "left": "issue:44493", + "right": "issue:46050", + "accept": false, + "reason": "Unexpected position-id key warnings are unrelated to NaNs in classification heads." + }, + { + "left": "issue:43653", + "right": "issue:44568", + "accept": false, + "reason": "Both involve special-token behavior, but one is a BigBird CI/decode issue and the other is mdeberta-v3 not adding BOS/EOS; distinct tokenizer bugs." + }, + { + "left": "issue:44843", + "right": "issue:44991", + "accept": false, + "reason": "Both are tokenizer-loading regressions, but the offline model_info bug and the est-roberta loading failure are different causes." + }, + { + "left": "issue:43756", + "right": "issue:46402", + "accept": false, + "reason": "Smollm3 RoPE-layer mismatch and Sam3 text-encoder weights not loaded are unrelated model-specific loading bugs." + }, + { + "left": "issue:29159", + "right": "issue:33357", + "accept": false, + "reason": "Slow/fast tokenizer inconsistency is unrelated to a MacOS bus error on a CLIP model." + }, + { + "left": "issue:43531", + "right": "issue:43742", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and MobileLLM key error are different model/configuration bugs." + }, + { + "left": "issue:43653", + "right": "issue:43927", + "accept": false, + "reason": "BigBird special-token registration and DiaConfig custom token-ID persistence are both token-related but not the same failure." + }, + { + "left": "issue:44464", + "right": "issue:44743", + "accept": false, + "reason": "Chunked generation with compiled forward and recurrent-state reset in modular_qwen3_5 are separate generation/cache issues." + }, + { + "left": "issue:44514", + "right": "issue:45290", + "accept": false, + "reason": "Both are apply_chat_template crashes, but one is batched padding=False behavior and the other is assistant tool-call messages with no content." + }, + { + "left": "issue:43550", + "right": "issue:43927", + "accept": false, + "reason": "torch.compile with SDPA in Bamba is unrelated to DiaConfig token ID loss after save/load." + }, + { + "left": "issue:43550", + "right": "issue:44361", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure is unrelated to MLukeTokenizer AttributeError." + }, + { + "left": "issue:44464", + "right": "issue:44977", + "accept": false, + "reason": "Both mention generation, but chunked compiled forward inconsistency and Qwen3.5 flash-attention generation are different concrete paths." + }, + { + "left": "issue:28282", + "right": "issue:35141", + "accept": false, + "reason": "Missing PyTorch import for AutoModel is unrelated to post_init reinitializing output embeddings after resize_token_embeddings()." + }, + { + "left": "issue:43643", + "right": "issue:44561", + "accept": false, + "reason": "Both involve trust_remote_code regressions, but one loses fields in AutoConfig and the other breaks due to removal of is_torch_fx_available." + }, + { + "left": "issue:45491", + "right": "issue:46002", + "accept": false, + "reason": "Both report NaNs, but one is Gemma3 sliding-window attention on GPU and the other is classification-head NaNs on checkout/init; 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": 20, + "review_unit_index": 23, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28282", + "issue:32090", + "issue:33666", + "issue:36010", + "issue:36331", + "issue:39692", + "issue:41720", + "issue:43262", + "issue:43295", + "issue:43452", + "issue:43493", + "issue:43504", + "issue:43525", + "issue:43550", + "issue:43572", + "issue:43577", + "issue:43606", + "issue:43688", + "issue:43720", + "issue:43742", + "issue:43749", + "issue:43825", + "issue:43927", + "issue:43994", + "issue:44361", + "issue:44451", + "issue:44464", + "issue:44466", + "issue:44479", + "issue:44554", + "issue:44568", + "issue:44610", + "issue:44625", + "issue:44821", + "issue:44849", + "issue:44964", + "issue:44977", + "issue:45200", + "issue:45229", + "issue:45357", + "issue:45405", + "issue:45440", + "issue:45588", + "issue:45823", + "issue:46050", + "issue:46192", + "issue:46209", + "issue:46489" + ], + "soft_pairs": [ + "issue:43504|issue:43720", + "issue:43504|issue:43550", + "issue:43504|issue:44610", + "issue:44466|issue:46050", + "issue:33666|issue:36010", + "issue:43742|issue:44821", + "issue:32090|issue:33666", + "issue:44977|issue:46209", + "issue:43577|issue:45588", + "issue:43688|issue:45440", + "issue:41720|issue:44464", + "issue:28282|issue:32090", + "issue:43720|issue:43927", + "issue:43262|issue:43295", + "issue:43606|issue:43927", + "issue:43452|issue:44451", + "issue:45823|issue:46050", + "issue:33666|issue:36331", + "issue:43994|issue:46192", + "issue:43525|issue:43572", + "issue:43720|issue:44361", + "issue:44849|issue:46209", + "issue:43493|issue:43825", + "issue:44964|issue:45405", + "issue:39692|issue:43493", + "issue:44479|issue:44625", + "issue:43577|issue:44554", + "issue:44849|issue:45357", + "issue:44568|issue:46489", + "issue:45200|issue:45229", + "issue:39692|issue:43295", + "issue:43688|issue:43749" + ], + "prepared_review_unit_hash": "bfac684d87a90facfdf5dc30e82e03e6c327229fbc0ad2804ea41b789a64dc2a", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37390, + "estimated_input_tokens": 9348, + "estimated_eval_tokens": 18952 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft-edge duplicate pairs look mergeable here; the issues mostly span different models, configs, or code paths, even when the symptom class is similar (loading, generation, device mapping, etc.).", + "confidence": 0.94, + "canonical_issue_reason": "If forced to anchor this heterogeneous set, issue 43452 is the most generic user-facing loading regression, but there is no truly canonical duplicate cluster to collapse around.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43452 is the best single representative only in the loose sense that it covers a broad from_pretrained loading failure; otherwise the set is too diverse for a strong global anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43504", + "right": "issue:43720", + "accept": false, + "reason": "Different bugs: legacy field loading for Beit vs BitNet packed-weight unpacking during accelerate loading." + }, + { + "left": "issue:43504", + "right": "issue:43550", + "accept": false, + "reason": "Different models and failures: Beit pretrained loading vs Bamba torch.compile/SDPA behavior." + }, + { + "left": "issue:43504", + "right": "issue:44610", + "accept": false, + "reason": "One is a legacy-field load error; the other is an OmDet processor size mismatch." + }, + { + "left": "issue:44466", + "right": "issue:46050", + "accept": false, + "reason": "Both involve serialization, but one is tied weights/device-dependent save behavior and the other is sinq quantized save/load breakage." + }, + { + "left": "issue:33666", + "right": "issue:36010", + "accept": false, + "reason": "Multi-GPU Qwen2-VL training issues are unrelated to the GenerationMixin import regression." + }, + { + "left": "issue:43742", + "right": "issue:44821", + "accept": false, + "reason": "A model key error and loading an AutoImageProcessor from URL are different load paths and failures." + }, + { + "left": "issue:32090", + "right": "issue:33666", + "accept": false, + "reason": "Trainer GPU broadcast NoneType error is a different distributed-training failure from Qwen2-VL multi-GPU training." + }, + { + "left": "issue:44977", + "right": "issue:46209", + "accept": false, + "reason": "Different runtime paths: Qwen3.5 flash-attention generation vs FP8 inference on B200." + }, + { + "left": "issue:43577", + "right": "issue:45588", + "accept": false, + "reason": "Blip2 dtype propagation and flash_attention s_aux=None crash are unrelated code paths." + }, + { + "left": "issue:43688", + "right": "issue:45440", + "accept": false, + "reason": "Aux-loss normalization in OLMoE/GPT Oss is not the same bug as DeepSeekV3MoE divergence." + }, + { + "left": "issue:41720", + "right": "issue:44464", + "accept": false, + "reason": "Auto device mapping cudaErrorAssert and chunked-generation compiled-forward inconsistency are distinct failures." + }, + { + "left": "issue:28282", + "right": "issue:32090", + "accept": false, + "reason": "PyTorch missing ImportError and Trainer _gpu_broadcast_one TypeError are unrelated setup vs distributed-runtime problems." + }, + { + "left": "issue:43720", + "right": "issue:43927", + "accept": false, + "reason": "BitNet packed-weight loading and DiaConfig token-ID persistence are unrelated bugs." + }, + { + "left": "issue:43262", + "right": "issue:43295", + "accept": false, + "reason": "Audio chat-template sampling-rate default and processor.tokenizer/image regression are different processor behaviors." + }, + { + "left": "issue:43606", + "right": "issue:43927", + "accept": false, + "reason": "CPU offload device mismatch and custom token ID loss are unrelated issues." + }, + { + "left": "issue:43452", + "right": "issue:44451", + "accept": false, + "reason": "General gguf/from_pretrained loading breakage is not the same as ScandiBERT-specific load failure." + }, + { + "left": "issue:45823", + "right": "issue:46050", + "accept": false, + "reason": "Gemma4 auto device-map device mismatch is unrelated to sinq quantization save/load corruption." + }, + { + "left": "issue:33666", + "right": "issue:36331", + "accept": false, + "reason": "Qwen2-VL multi-GPU training and custom trainer compute_loss signature mismatch are different regressions." + }, + { + "left": "issue:43994", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 wrong outputs and return_tensors being ignored are different API bugs." + }, + { + "left": "issue:43525", + "right": "issue:43572", + "accept": false, + "reason": "Similar theme of missing config fields, but different models and different missing attributes." + }, + { + "left": "issue:43720", + "right": "issue:44361", + "accept": false, + "reason": "BitNet packed-weight loading is unrelated to MLukeTokenizer AttributeError on tasks." + }, + { + "left": "issue:44849", + "right": "issue:46209", + "accept": false, + "reason": "Qwen3.5 hidden-state output bug and FP8 B200 inference failure are unrelated." + }, + { + "left": "issue:43493", + "right": "issue:43825", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and pipeline translation-message wording are not the same bug." + }, + { + "left": "issue:44964", + "right": "issue:45405", + "accept": false, + "reason": "Phi-4 multimodal load failure and MIN_PEFT_VERSION packaging issue are unrelated." + }, + { + "left": "issue:39692", + "right": "issue:43493", + "accept": false, + "reason": "Docs-example errors and core SigLIP2 implementation discrepancy are different issues." + }, + { + "left": "issue:44479", + "right": "issue:44625", + "accept": false, + "reason": "Video-input regression across Qwen models is separate from Qwen3.5 num_labels propagation." + }, + { + "left": "issue:43577", + "right": "issue:44554", + "accept": false, + "reason": "Blip2 dtype loading and MPS attention correctness differ in subsystem and symptom." + }, + { + "left": "issue:44849", + "right": "issue:45357", + "accept": false, + "reason": "Hidden-states output bug and incorrect visual-encoder save_pretrained keys are distinct Qwen3.5 regressions." + }, + { + "left": "issue:44568", + "right": "issue:46489", + "accept": false, + "reason": "Different tokenizer families and failures; mdeberta special tokens vs DeepSeek-Coder v1 output regression." + }, + { + "left": "issue:45200", + "right": "issue:45229", + "accept": false, + "reason": "Gemma4 text-only token-type defaults and Gemma4 31B multi-GPU OOM are unrelated problems." + }, + { + "left": "issue:39692", + "right": "issue:43295", + "accept": false, + "reason": "SigLIP2 documentation errors and processor.tokenizer/image regression do not share the same code-path bug." + }, + { + "left": "issue:43688", + "right": "issue:43749", + "accept": false, + "reason": "Aux-loss normalization and FSDP CPU RAM efficient loading are different training/loading 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": 20, + "review_unit_index": 24, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:33357", + "issue:33666", + "issue:43257", + "issue:43278", + "issue:43299", + "issue:43334", + "issue:43493", + "issue:43504", + "issue:43540", + "issue:43572", + "issue:43575", + "issue:43606", + "issue:43646", + "issue:43653", + "issue:43688", + "issue:43824", + "issue:43844", + "issue:43927", + "issue:44112", + "issue:44291", + "issue:44361", + "issue:44423", + "issue:44464", + "issue:44466", + "issue:44488", + "issue:44493", + "issue:44554", + "issue:44568", + "issue:44589", + "issue:44625", + "issue:44734", + "issue:44792", + "issue:44843", + "issue:44977", + "issue:45229", + "issue:45405", + "issue:45520", + "issue:45588", + "issue:45701", + "issue:45704", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:46050", + "issue:46285", + "issue:46365", + "issue:46424" + ], + "soft_pairs": [ + "issue:44488|issue:44843", + "issue:43540|issue:44977", + "issue:44568|issue:45812", + "issue:44464|issue:46424", + "issue:43493|issue:43844", + "issue:43646|issue:44466", + "issue:43334|issue:43572", + "issue:43824|issue:44589", + "issue:33357|issue:33666", + "issue:44589|issue:45520", + "issue:43688|issue:45704", + "issue:43334|issue:45759", + "issue:43278|issue:43844", + "issue:43504|issue:43606", + "issue:43927|issue:44361", + "issue:45229|issue:45823", + "issue:43572|issue:44291", + "issue:44792|issue:45405", + "issue:44493|issue:46285", + "issue:43257|issue:43299", + "issue:43575|issue:45405", + "issue:43504|issue:43653", + "issue:43606|issue:44112", + "issue:44625|issue:45759", + "issue:44554|issue:46365", + "issue:43688|issue:45701", + "issue:45588|issue:46050", + "issue:44423|issue:44734" + ], + "prepared_review_unit_hash": "a9bba81ac26be61bd4560fb52d8474e5b1a38b42a6683c09078ea55b4988dfbb", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35685, + "estimated_input_tokens": 8922, + "estimated_eval_tokens": 18100 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is very heterogeneous: most items are isolated Transformers bugs across different models, configs, and failure modes. The soft-similarity links are all false positives; none look like the same underlying bug or fixable in one PR.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "No single issue cleanly represents the cluster because the items span unrelated bugs; if forced, the closest broad umbrella is still not a true canonical duplicate target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44488", + "right": "issue:44843", + "accept": false, + "reason": "Both are loading-related, but one is a specific model load failure and the other is an offline-mode hub call regression; different code paths and fixes." + }, + { + "left": "issue:43540", + "right": "issue:44977", + "accept": false, + "reason": "Different Qwen3 variants and symptoms: one is a video-input ValueError in Qwen3OmniMoe, the other is flash-attention generation quality in Qwen3.5." + }, + { + "left": "issue:44568", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different model families and different tokenization behavior." + }, + { + "left": "issue:44464", + "right": "issue:46424", + "accept": false, + "reason": "Chunked generation compiled-forward inconsistency is unrelated to cache max length handling for quantized cache." + }, + { + "left": "issue:43493", + "right": "issue:43844", + "accept": false, + "reason": "One is a model-implementation discrepancy vs JAX; the other is a DeepSpeed training instability issue." + }, + { + "left": "issue:43646", + "right": "issue:44466", + "accept": false, + "reason": "Generic custom model initialization breakage is not the same bug as device-dependent lm_head serialization inconsistency." + }, + { + "left": "issue:43334", + "right": "issue:43572", + "accept": false, + "reason": "Both mention missing config fields, but they are different fields on different model classes with distinct failure modes." + }, + { + "left": "issue:43824", + "right": "issue:44589", + "accept": false, + "reason": "Importing a missing VL class and failing on Float8 storage lookup are unrelated failures." + }, + { + "left": "issue:33357", + "right": "issue:33666", + "accept": false, + "reason": "MacOS CLIP bus error and Qwen2-VL multi-GPU training are entirely different problems." + }, + { + "left": "issue:44589", + "right": "issue:45520", + "accept": false, + "reason": "Float8 storage lookup error is unrelated to a Python 3.13 flash_attn import-utils KeyError." + }, + { + "left": "issue:43688", + "right": "issue:45704", + "accept": false, + "reason": "Aux-loss normalization in MoE models is unrelated to a T5 apex RMSNorm memory leak." + }, + { + "left": "issue:43334", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3-VL pad_token_id loss and composite Qwen 3.5/3.6 text_config unwrapping are different configuration bugs." + }, + { + "left": "issue:43278", + "right": "issue:43844", + "accept": false, + "reason": "Embedding dtype drift during evaluate is unrelated to the ZeRO-3 gradient increase issue." + }, + { + "left": "issue:43504", + "right": "issue:43606", + "accept": false, + "reason": "Legacy-field model loading failure and CPU-offload device mismatch do not share a concrete code path." + }, + { + "left": "issue:43927", + "right": "issue:44361", + "accept": false, + "reason": "DiaConfig token ID persistence and MLukeTokenizer task AttributeError are different model/tokenizer issues." + }, + { + "left": "issue:45229", + "right": "issue:45823", + "accept": false, + "reason": "Both are Gemma4, but one is OOM under multi-GPU inference while the other is a forward-time device mismatch." + }, + { + "left": "issue:43572", + "right": "issue:44291", + "accept": false, + "reason": "Both are v5 compatibility issues, but they concern different APIs: missing pad_token_idx vs unexpected _is_hf_initialized in init_empty_weights." + }, + { + "left": "issue:44792", + "right": "issue:45405", + "accept": false, + "reason": "A failing Janus image-generation test is unrelated to a PEFT version pin bump." + }, + { + "left": "issue:44493", + "right": "issue:46285", + "accept": false, + "reason": "Unexpected position-id keys in many models is not the same as the CLIP text-model structure change." + }, + { + "left": "issue:43257", + "right": "issue:43299", + "accept": false, + "reason": "Qwen3 MoE weight conversion with accelerate/deepspeed and Qwen3VL MoE load failure are related only by family name, not the same bug." + }, + { + "left": "issue:43575", + "right": "issue:45405", + "accept": false, + "reason": "Tensor-parallel OOM on a Qwen2 model is unrelated to an unreleased dependency-version bump." + }, + { + "left": "issue:43504", + "right": "issue:43653", + "accept": false, + "reason": "Beit pretrained load failure and BigBirdTokenizer special-token registration are different subsystems and bugs." + }, + { + "left": "issue:43606", + "right": "issue:44112", + "accept": false, + "reason": "CPU offload device mismatch and a stale GraniteSpeech CI test are not the same underlying issue." + }, + { + "left": "issue:44625", + "right": "issue:45759", + "accept": false, + "reason": "Qwen3.5 num_labels propagation and composite Qwen config unwrapping are distinct config-serialization problems." + }, + { + "left": "issue:44554", + "right": "issue:46365", + "accept": false, + "reason": "An MPS attention correctness bug is unrelated to Trainer import failure caused by datasets installation." + }, + { + "left": "issue:43688", + "right": "issue:45701", + "accept": false, + "reason": "MoE auxiliary-loss normalization and a broad tokenization regression report are not the same concrete bug." + }, + { + "left": "issue:45588", + "right": "issue:46050", + "accept": false, + "reason": "Flash-attention sink-less-model crash and sinq save/load quantization breakage are unrelated." + }, + { + "left": "issue:44423", + "right": "issue:44734", + "accept": false, + "reason": "Both involve serve, but one crashes on multimodal continuous batching and the other on KV-cache continuation indexing." + } + ] + }, + "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": 20, + "review_unit_index": 25, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:15354", + "issue:28218", + "issue:29159", + "issue:29942", + "issue:30333", + "issue:32090", + "issue:33357", + "issue:34567", + "issue:38175", + "issue:39692", + "issue:43065", + "issue:43232", + "issue:43278", + "issue:43295", + "issue:43450", + "issue:43493", + "issue:43540", + "issue:43550", + "issue:43572", + "issue:43688", + "issue:43701", + "issue:43844", + "issue:43873", + "issue:43994", + "issue:44186", + "issue:44466", + "issue:44493", + "issue:44521", + "issue:44554", + "issue:44610", + "issue:44849", + "issue:44964", + "issue:44987", + "issue:45072", + "issue:45084", + "issue:45127", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45356", + "issue:45406", + "issue:45446", + "issue:45520", + "issue:45823", + "issue:46050", + "issue:46121", + "issue:46192", + "issue:46369" + ], + "soft_pairs": [ + "issue:29942|issue:30333", + "issue:43844|issue:44964", + "issue:43994|issue:44493", + "issue:43688|issue:43873", + "issue:29159|issue:32090", + "issue:28218|issue:34567", + "issue:45084|issue:45520", + "issue:44554|issue:45446", + "issue:15354|issue:38175", + "issue:43540|issue:45200", + "issue:43065|issue:46369", + "issue:43278|issue:43701", + "issue:43873|issue:46050", + "issue:45072|issue:45198", + "issue:45200|issue:45823", + "issue:44186|issue:46192", + "issue:44521|issue:45290", + "issue:43232|issue:44849", + "issue:43493|issue:43701", + "issue:45406|issue:46121", + "issue:45823|issue:46192", + "issue:43295|issue:43450", + "issue:28218|issue:33357", + "issue:43295|issue:44610", + "issue:44554|issue:45356", + "issue:43295|issue:43873", + "issue:43572|issue:44987", + "issue:39692|issue:46192", + "issue:43550|issue:43844", + "issue:44466|issue:45127" + ], + "prepared_review_unit_hash": "708fa07163da345887dd3c38399925f8b4976c4ec60fa3ecd489af3102e711f2", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36687, + "estimated_input_tokens": 9172, + "estimated_eval_tokens": 18600 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "No soft edge pair looks like a true duplicate; the set is a grab-bag of unrelated tokenizer, processor, quantization, generation, and training issues rather than one coherent bug cluster.", + "confidence": 0.94, + "canonical_issue_reason": "No single canonical issue: the items are heterogeneous and the soft-pair similarities are superficial.", + "canonical_pr_reason": null, + "best_issue_reason": "No issue is a convincing cluster representative because none of the candidates share the same underlying defect.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:29942", + "right": "issue:30333", + "accept": false, + "reason": "Different failures in different areas: FlashAttention test regressions vs MLflow reporting on failed jobs." + }, + { + "left": "issue:43844", + "right": "issue:44964", + "accept": false, + "reason": "Unrelated bugs: ZeRO-3 gradient growth vs model loading failure for Phi-4 multimodal." + }, + { + "left": "issue:43994", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 output quality issue vs position-id key mismatch; different code paths and symptoms." + }, + { + "left": "issue:43688", + "right": "issue:43873", + "accept": false, + "reason": "Aux-loss normalization bug is unrelated to quantization/offloading behavior." + }, + { + "left": "issue:29159", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer slow/fast inconsistency vs Trainer GPU broadcast TypeError are unrelated." + }, + { + "left": "issue:28218", + "right": "issue:34567", + "accept": false, + "reason": "Added-token spacing bug is unrelated to TrainerState token counting." + }, + { + "left": "issue:45084", + "right": "issue:45520", + "accept": false, + "reason": "Template-node compilation error and flash_attn import KeyError are separate issues." + }, + { + "left": "issue:44554", + "right": "issue:45446", + "accept": false, + "reason": "Both mention attention-related code, but one is an MPS correctness bug and the other is a PyTorch version guard/import check." + }, + { + "left": "issue:15354", + "right": "issue:38175", + "accept": false, + "reason": "TorchScript GeneratorExp export failure vs SigLIP2 zero probabilities are unrelated." + }, + { + "left": "issue:43540", + "right": "issue:45200", + "accept": false, + "reason": "Video input processing error in Qwen3OmniMoe vs Gemma4 multimodal token-type defaulting." + }, + { + "left": "issue:43065", + "right": "issue:46369", + "accept": false, + "reason": "Dummy Conv2d in SAM3 pixel decoder is not the same as memory overflow in sam3-litetext." + }, + { + "left": "issue:43278", + "right": "issue:43701", + "accept": false, + "reason": "BF16/FP32 dtype drift during evaluate is unrelated to resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43873", + "right": "issue:46050", + "accept": false, + "reason": "Quantization/offloading symptoms differ: offloading not working vs broken save/load round-trip after sinq quantization." + }, + { + "left": "issue:45072", + "right": "issue:45198", + "accept": false, + "reason": "bfloat16 dtype mismatches in inference are unrelated to Wav2Vec2 save/tokenization failures." + }, + { + "left": "issue:44521", + "right": "issue:45290", + "accept": false, + "reason": "Both touch apply_chat_template, but one is zero assistant masks for multimodal inputs and the other is a crash on tool-call assistant messages; different edge cases." + }, + { + "left": "issue:43232", + "right": "issue:44849", + "accept": false, + "reason": "Generation kwargs update after sync_gpus is unrelated to Qwen3.5 output_hidden_states bug." + }, + { + "left": "issue:43493", + "right": "issue:43701", + "accept": false, + "reason": "SigLIP2 HF-vs-JAX discrepancy is unrelated to checkpoint resume key mismatch." + }, + { + "left": "issue:45406", + "right": "issue:46121", + "accept": false, + "reason": "Gemma4Processor _tokenizer crash and rope-params list TypeError are different bugs." + }, + { + "left": "issue:45823", + "right": "issue:46192", + "accept": false, + "reason": "Device-map forward mismatch is unrelated to return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:43295", + "right": "issue:43450", + "accept": false, + "reason": "Processor.tokenizer regression is unrelated to batched video processor shape errors." + }, + { + "left": "issue:28218", + "right": "issue:33357", + "accept": false, + "reason": "Added-token spacing bug is unrelated to MacOS bus error in CLIP model loading." + }, + { + "left": "issue:43295", + "right": "issue:44610", + "accept": false, + "reason": "Custom processor/tokenizer regression is not the same as OmDet-Turbo\u2019s 640px-vs-224px preprocessing mismatch." + }, + { + "left": "issue:44554", + "right": "issue:45356", + "accept": false, + "reason": "Different attention/tokenizer problems: MPS attention correctness vs Kimi-K2.5 tokenizer regression." + }, + { + "left": "issue:43295", + "right": "issue:43873", + "accept": false, + "reason": "Processor/tokenizer regression is unrelated to quantization offloading behavior." + }, + { + "left": "issue:43572", + "right": "issue:44987", + "accept": false, + "reason": "Missing pad_token_idx in StableLmConfig is not the same as loading failure for physical-intelligence/fast." + }, + { + "left": "issue:39692", + "right": "issue:46192", + "accept": false, + "reason": "SigLIP2 doc example errors are unrelated to return_tensors being ignored with text_kwargs." + }, + { + "left": "issue:43550", + "right": "issue:43844", + "accept": false, + "reason": "torch.compile+SDPA failure for Bamba-9B-v2 is unrelated to ZeRO-3 gradient blow-up." + }, + { + "left": "issue:44466", + "right": "issue:45127", + "accept": false, + "reason": "Both involve weights, but one is inconsistent serialization and the other is LoRA merge collapse with extended vocab." + } + ] + }, + "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": 20, + "review_unit_index": 26, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:37428", + "issue:38175", + "issue:42490", + "issue:42915", + "issue:43262", + "issue:43278", + "issue:43295", + "issue:43334", + "issue:43450", + "issue:43493", + "issue:43525", + "issue:43526", + "issue:43572", + "issue:43575", + "issue:43653", + "issue:43696", + "issue:43819", + "issue:43824", + "issue:43827", + "issue:43856", + "issue:43901", + "issue:43950", + "issue:43957", + "issue:43975", + "issue:44112", + "issue:44442", + "issue:44514", + "issue:44521", + "issue:44561", + "issue:44610", + "issue:44857", + "issue:44898", + "issue:45083", + "issue:45198", + "issue:45200", + "issue:45290", + "issue:45325", + "issue:45381", + "issue:45412", + "issue:45520", + "issue:45759", + "issue:46170", + "issue:46178", + "issue:46192", + "issue:46209", + "issue:46234", + "issue:46285", + "issue:46326" + ], + "soft_pairs": [ + "issue:43572|issue:43957", + "issue:44898|issue:45412", + "issue:43950|issue:46285", + "issue:45520|issue:46178", + "issue:45381|issue:46209", + "issue:45200|issue:45325", + "issue:43450|issue:43575", + "issue:37428|issue:43957", + "issue:43334|issue:43525", + "issue:43824|issue:43975", + "issue:38175|issue:43493", + "issue:43653|issue:44112", + "issue:42490|issue:43696", + "issue:43278|issue:43856", + "issue:43526|issue:46192", + "issue:42915|issue:43824", + "issue:43295|issue:43493", + "issue:44442|issue:45198", + "issue:37428|issue:43824", + "issue:38175|issue:45412", + "issue:45520|issue:46170", + "issue:43262|issue:43450", + "issue:44610|issue:44857", + "issue:44561|issue:46234", + "issue:45759|issue:46285", + "issue:43819|issue:44112", + "issue:43827|issue:43901", + "issue:45290|issue:46326", + "issue:44514|issue:44521", + "issue:43819|issue:45083" + ], + "prepared_review_unit_hash": "7b505cebe9dc030015e3494c0820482268fd2e3afc67eb2964d41f835ef72f86", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36973, + "estimated_input_tokens": 9244, + "estimated_eval_tokens": 18744 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is not a single duplicate cluster; it\u2019s a grab bag of mostly unrelated Transformers regressions, with only a loose multimodal/chat-template subtheme. None of the soft pairs are close enough to merge as true duplicates, so I\u2019d treat them as distinct issues and use issue 44521 only as the broadest representative of the chat-template subgroup.", + "confidence": 0.89, + "canonical_issue_reason": "If a representative issue is needed, 44521 is the broadest and most active multimodal apply_chat_template regression in this set, but the cluster overall is not a true duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "44521 is the most central issue among the few related items: it has the broadest multimodal chat-template scope, active discussion, and references, making it the best stand-in for the loose subgroup.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43572", + "right": "issue:43957", + "accept": false, + "reason": "Both are v5 loading regressions, but StableLm missing pad_token_idx and meta-device loading are different root causes." + }, + { + "left": "issue:44898", + "right": "issue:45412", + "accept": false, + "reason": "Different bugs: Perceiver non-default-resolution interpolation vs RT-DETR memory not being released." + }, + { + "left": "issue:43950", + "right": "issue:46285", + "accept": false, + "reason": "Both involve model-loading structure changes, but buffer persistence corruption and CLIP text-model layout inconsistency are not the same defect." + }, + { + "left": "issue:45520", + "right": "issue:46178", + "accept": false, + "reason": "Separate import/backends failures: flash_attn lookup on Python 3.13 vs tensorflow_text not in BACKENDS_MAPPING." + }, + { + "left": "issue:45381", + "right": "issue:46209", + "accept": false, + "reason": "Unrelated code paths: Qwen2.5-VL rope/position ids vs FP8 inference failure on B200." + }, + { + "left": "issue:45200", + "right": "issue:45325", + "accept": false, + "reason": "Both touch multimodal defaults, but Gemma 4 mm_token_type_ids and Qwen2.5-VL temporal position ids are distinct issues." + }, + { + "left": "issue:43450", + "right": "issue:43575", + "accept": false, + "reason": "Video processor batch-shape bug and tensor-parallel OOM are unrelated." + }, + { + "left": "issue:37428", + "right": "issue:43957", + "accept": false, + "reason": "ImportError for a flash-attention helper is not the same as meta-device model loading breakage." + }, + { + "left": "issue:43334", + "right": "issue:43525", + "accept": false, + "reason": "Same symptom name, different models/config classes; no evidence of one underlying bug." + }, + { + "left": "issue:43824", + "right": "issue:43975", + "accept": false, + "reason": "Import error for Qwen2.5-VL class vs DeepSeek detokenization regression are different problems." + }, + { + "left": "issue:38175", + "right": "issue:43493", + "accept": false, + "reason": "Both are SigLIP2-related, but one is zero probabilities and the other is a broader HF-vs-JAX discrepancy; not enough to treat as one concrete bug." + }, + { + "left": "issue:43653", + "right": "issue:44112", + "accept": false, + "reason": "Tokenizer special-token registration bug vs GraniteSpeech CI test flake are unrelated." + }, + { + "left": "issue:42490", + "right": "issue:43696", + "accept": false, + "reason": "Local-path import warning and GPT-oss OOM are unrelated failure modes." + }, + { + "left": "issue:43278", + "right": "issue:43856", + "accept": false, + "reason": "BF16-to-FP32 dtype drift during eval is not the same as MoE training memory inefficiency." + }, + { + "left": "issue:43526", + "right": "issue:46192", + "accept": false, + "reason": "Both are processor argument-handling bugs, but reduce_labels truncation and return_tensors/text_kwargs precedence are different defects." + }, + { + "left": "issue:42915", + "right": "issue:43824", + "accept": false, + "reason": "Qwen3MoE FP8 config failure and Qwen2.5-VL import error are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:43493", + "accept": false, + "reason": "Custom processor.tokenizer regression and SigLIP2 implementation discrepancy do not share the same code-path." + }, + { + "left": "issue:44442", + "right": "issue:45198", + "accept": false, + "reason": "Both are speech tokenization issues, but they affect different tokenizer families and different save/load paths." + }, + { + "left": "issue:37428", + "right": "issue:43824", + "accept": false, + "reason": "Different import errors in different modules/classes." + }, + { + "left": "issue:38175", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 output anomaly and RT-DETR memory retention are unrelated." + }, + { + "left": "issue:45520", + "right": "issue:46170", + "accept": false, + "reason": "Both are import-related, but the Windows ROCm torch.distributed issue is a different dependency problem." + }, + { + "left": "issue:43262", + "right": "issue:43450", + "accept": false, + "reason": "Audio chat-template sampling-rate default and video batched-shape bug are different processor regressions." + }, + { + "left": "issue:44610", + "right": "issue:44857", + "accept": false, + "reason": "Different model/task bugs: OmDet processor sizing vs LwDetr AMP crash." + }, + { + "left": "issue:44561", + "right": "issue:46234", + "accept": false, + "reason": "Both are missing helper symbols in import_utils, but they are different functions with different downstream breakages." + }, + { + "left": "issue:45759", + "right": "issue:46285", + "accept": false, + "reason": "Both involve composite-model structure, but Qwen multimodal config unwrapping and CLIP text-model layout inconsistency are not the same bug." + }, + { + "left": "issue:43819", + "right": "issue:44112", + "accept": false, + "reason": "Unrelated: DAC latent reconstruction mismatch vs GraniteSpeech CI device override." + }, + { + "left": "issue:43827", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs regressions, but they cover different pipelines/APIs and are not duplicates." + }, + { + "left": "issue:45290", + "right": "issue:46326", + "accept": false, + "reason": "Both concern chat-template edge cases, but tool-call/no-content crashes and mistral_common continue_final_message behavior are distinct." + }, + { + "left": "issue:44514", + "right": "issue:44521", + "accept": false, + "reason": "Same area, but one is a batched padding=False crash and the other is all-zero assistant masks; different concrete failures." + }, + { + "left": "issue:43819", + "right": "issue:45083", + "accept": false, + "reason": "Completely different areas: DAC from_latents vs qwen3_omni_moe helper output lengths." + } + ] + }, + "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": 20, + "review_unit_index": 27, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28218", + "issue:29942", + "issue:32090", + "issue:35141", + "issue:37428", + "issue:39692", + "issue:41720", + "issue:42796", + "issue:43278", + "issue:43335", + "issue:43388", + "issue:43425", + "issue:43504", + "issue:43526", + "issue:43550", + "issue:43606", + "issue:43611", + "issue:43653", + "issue:43720", + "issue:43819", + "issue:43825", + "issue:44112", + "issue:44206", + "issue:44351", + "issue:44448", + "issue:44493", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44857", + "issue:45083", + "issue:45127", + "issue:45200", + "issue:45245", + "issue:45290", + "issue:45310", + "issue:45381", + "issue:45459", + "issue:45478", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45823", + "issue:45859", + "issue:46209", + "issue:46234", + "issue:46402" + ], + "soft_pairs": [ + "issue:41720|issue:46209", + "issue:44857|issue:45593", + "issue:43335|issue:43550", + "issue:43611|issue:46402", + "issue:44206|issue:45584", + "issue:43606|issue:45083", + "issue:37428|issue:43425", + "issue:42796|issue:46209", + "issue:44792|issue:45563", + "issue:44448|issue:44743", + "issue:43335|issue:43653", + "issue:45200|issue:45588", + "issue:45127|issue:45245", + "issue:43550|issue:44112", + "issue:44855|issue:45083", + "issue:37428|issue:44351", + "issue:43335|issue:43504", + "issue:43819|issue:44493", + "issue:43278|issue:43825", + "issue:44206|issue:45823", + "issue:29942|issue:32090", + "issue:39692|issue:43526", + "issue:45588|issue:45859", + "issue:43335|issue:45083", + "issue:43720|issue:44112", + "issue:43388|issue:43526", + "issue:45459|issue:46234", + "issue:45310|issue:45478", + "issue:43335|issue:44112", + "issue:43335|issue:45593", + "issue:45290|issue:45381", + "issue:39692|issue:44493", + "issue:28218|issue:35141", + "issue:43606|issue:44857" + ], + "prepared_review_unit_hash": "9293f69383012f7bb181a4ecc6a2cd929c2a827e97892b8039eafbc04283f245", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 34, + "serialized_chars": 38089, + "estimated_input_tokens": 9523, + "estimated_eval_tokens": 19302 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is mostly a collection of unrelated issue reports. Only one soft pair is a clear duplicate: the two identical Qwen3.5 Moe `from_pretrained` errors (45310/45478).", + "confidence": 0.97, + "canonical_issue_reason": "issue:45310 is the clearest canonical issue: it is a narrowly scoped, reproducible bug report and has an exact duplicate in issue:45478 with the same title and failure mode.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45310 is the best overall issue candidate because it is specific, well-scoped, and has the strongest duplicate signal in the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41720", + "right": "issue:46209", + "accept": false, + "reason": "Both are model/runtime failures, but they concern different hardware and failure modes (cuda assert on A800 vs FP8 inference on B200)." + }, + { + "left": "issue:44857", + "right": "issue:45593", + "accept": false, + "reason": "Different models and bugs: AMP/CUDA crash in LwDetrImageLoss vs missing auxiliary losses in D-FINE." + }, + { + "left": "issue:43335", + "right": "issue:43550", + "accept": false, + "reason": "Unrelated code paths: SwitchTransformers config layer creation vs Bamba torch.compile/SDPA failure." + }, + { + "left": "issue:43611", + "right": "issue:46402", + "accept": false, + "reason": "Both involve model loading, but the underlying bugs are different: `base_model_prefix` loading regression vs Sam3 text encoder weights not loaded." + }, + { + "left": "issue:44206", + "right": "issue:45584", + "accept": false, + "reason": "Different affected features and failures: LasrFeatureExtractor unsupported arg vs Whisper empty transcription after token alignment." + }, + { + "left": "issue:43606", + "right": "issue:45083", + "accept": false, + "reason": "Device mismatch with CPU offload is unrelated to the qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:37428", + "right": "issue:43425", + "accept": false, + "reason": "Missing internal flash-attention symbol is not the same as a Torch 2.10 compatibility issue." + }, + { + "left": "issue:42796", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer BPE marker handling and FP8 inference on B200 are distinct problems." + }, + { + "left": "issue:44792", + "right": "issue:45563", + "accept": false, + "reason": "A failing image-generation test is unrelated to a stale warning in paged generate()." + }, + { + "left": "issue:44448", + "right": "issue:44743", + "accept": false, + "reason": "Different model behaviors: Pegasus output regression vs recurrent states resetting with cache in qwen3 modular code." + }, + { + "left": "issue:43335", + "right": "issue:43653", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug and BigBirdTokenizer special-token bug are separate components and failure modes." + }, + { + "left": "issue:45200", + "right": "issue:45588", + "accept": false, + "reason": "Gemma4 text-only mm token IDs issue is unrelated to the flash_attention `s_aux=None` crash." + }, + { + "left": "issue:45127", + "right": "issue:45245", + "accept": false, + "reason": "LoRA/extended vocabulary collapse is unrelated to the category-count runtime error." + }, + { + "left": "issue:43550", + "right": "issue:44112", + "accept": false, + "reason": "Bamba torch.compile/SDPA failure and GraniteSpeech CI device-override flake are unrelated." + }, + { + "left": "issue:44855", + "right": "issue:45083", + "accept": false, + "reason": "Python 3.13 import/torch.jit parsing issue is unrelated to qwen3_omni_moe feature-length helper behavior." + }, + { + "left": "issue:37428", + "right": "issue:44351", + "accept": false, + "reason": "Missing flash-attention helper import and missing `HybridCache` import are distinct API regressions." + }, + { + "left": "issue:43335", + "right": "issue:43504", + "accept": false, + "reason": "SwitchTransformers sparse-layer creation bug is unrelated to Beit semantic segmentation pretrained loading with a legacy field." + }, + { + "left": "issue:43819", + "right": "issue:44493", + "accept": false, + "reason": "DAC latent/forward mismatch is unrelated to widespread unexpected position-id keys." + }, + { + "left": "issue:43278", + "right": "issue:43825", + "accept": false, + "reason": "Embedding dtype drift between train/eval is unrelated to a pipeline error-message regression." + }, + { + "left": "issue:44206", + "right": "issue:45823", + "accept": false, + "reason": "Different models and failure modes: LasrFeatureExtractor crash vs Gemma4 PLE device mismatch." + }, + { + "left": "issue:29942", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 test failures and Trainer `_gpu_broadcast_one` NoneType errors are separate issues." + }, + { + "left": "issue:39692", + "right": "issue:43526", + "accept": false, + "reason": "Documentation example errors for SigLIP2 are unrelated to BeitImageProcessorFast label reduction behavior." + }, + { + "left": "issue:45588", + "right": "issue:45859", + "accept": false, + "reason": "Both touch Qwen3-related code, but one is a flash-attention sink crash and the other is CPU-offload incompatibility in rotary embedding; not the same bug." + }, + { + "left": "issue:43335", + "right": "issue:45083", + "accept": false, + "reason": "SwitchTransformers config bug and qwen3_omni_moe helper behavior are unrelated." + }, + { + "left": "issue:43720", + "right": "issue:44112", + "accept": false, + "reason": "BitNet packed-weight loading and GraniteSpeech stale CI override are unrelated." + }, + { + "left": "issue:43388", + "right": "issue:43526", + "accept": false, + "reason": "Metric-gathering label truncation and Beit label reduction bug affect different subsystems and data flows." + }, + { + "left": "issue:45459", + "right": "issue:46234", + "accept": false, + "reason": "Tokenizer error masking with missing protobuf and a missing import-utils availability helper are unrelated." + }, + { + "left": "issue:45310", + "right": "issue:45478", + "accept": true, + "reason": "Exact duplicate reports: same title, same Qwen3.5 Moe `from_pretrained` failure, same scope." + }, + { + "left": "issue:43335", + "right": "issue:44112", + "accept": false, + "reason": "SwitchTransformers config and GraniteSpeech CI test flake are unrelated." + }, + { + "left": "issue:43335", + "right": "issue:45593", + "accept": false, + "reason": "SwitchTransformers sparse-layer config bug is unrelated to D-FINE auxiliary-loss handling." + }, + { + "left": "issue:45290", + "right": "issue:45381", + "accept": false, + "reason": "Chat template tool-call crash and qwen2.5-vl video position-id bug are different code paths." + }, + { + "left": "issue:39692", + "right": "issue:44493", + "accept": false, + "reason": "SigLIP2 docs/example errors are unrelated to unexpected position-id keys across models." + }, + { + "left": "issue:28218", + "right": "issue:35141", + "accept": false, + "reason": "Tokenizer added-space behavior and output-embedding reinitialization are unrelated bugs." + }, + { + "left": "issue:43606", + "right": "issue:44857", + "accept": false, + "reason": "CPU offload device mismatch and float16 AMP crash are different model/runtime failures." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The 45310/45478 duplicate call is supported by the identical titles, and the other soft pairs are rejected without 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": 20, + "review_unit_index": 28, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:28282", + "issue:29127", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:34689", + "issue:35141", + "issue:36010", + "issue:37428", + "issue:38175", + "issue:40444", + "issue:42490", + "issue:43278", + "issue:43381", + "issue:43412", + "issue:43526", + "issue:43550", + "issue:43575", + "issue:43701", + "issue:43746", + "issue:43854", + "issue:43927", + "issue:44360", + "issue:44448", + "issue:44466", + "issue:44493", + "issue:44561", + "issue:44568", + "issue:44779", + "issue:44855", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45083", + "issue:45198", + "issue:45229", + "issue:45372", + "issue:45491", + "issue:45563", + "issue:45588", + "issue:45698", + "issue:45803", + "issue:45823", + "issue:45920", + "issue:46192", + "issue:46209", + "issue:46399" + ], + "soft_pairs": [ + "issue:44448|issue:44779", + "issue:42490|issue:43854", + "issue:44360|issue:44561", + "issue:44857|issue:45072", + "issue:38175|issue:44493", + "issue:45920|issue:46192", + "issue:34689|issue:35141", + "issue:43701|issue:45083", + "issue:44977|issue:45229", + "issue:44448|issue:44977", + "issue:44855|issue:46399", + "issue:29127|issue:32090", + "issue:43550|issue:45083", + "issue:43927|issue:45698", + "issue:42490|issue:43575", + "issue:40444|issue:43526", + "issue:29942|issue:35141", + "issue:44561|issue:44779", + "issue:27957|issue:34689", + "issue:44857|issue:45588", + "issue:44977|issue:45563", + "issue:44568|issue:45491", + "issue:28282|issue:33357", + "issue:43278|issue:43381", + "issue:43412|issue:45588", + "issue:45372|issue:45588", + "issue:45588|issue:45823", + "issue:36010|issue:37428", + "issue:43746|issue:46209", + "issue:44448|issue:44466", + "issue:45198|issue:45803" + ], + "prepared_review_unit_hash": "c1f00dc15778c440f8311ff1cba0a621fba6b42cb99c0b90310b6d1e8825f269", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 31, + "serialized_chars": 37214, + "estimated_input_tokens": 9304, + "estimated_eval_tokens": 18864 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These issues are mostly unrelated. The soft pairs share broad surface area (tokenizers, loading, flash-attn, mixed precision, etc.) but not the same concrete bug or fix path, so none of the candidate duplicates should be merged.", + "confidence": 0.94, + "canonical_issue_reason": "issue:45920 is the broadest, most central user-facing regression here: it affects AutoTokenizer across multiple models and is likely to subsume related tokenizer reports better than the narrower issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45920 is the strongest representative issue because it is open, impacts several model families, and describes a high-level regression with clear user impact.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44448", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer/output regressions, but they affect different models and likely different code paths." + }, + { + "left": "issue:42490", + "right": "issue:43854", + "accept": false, + "reason": "A local-path import warning is not the same bug as failing to load a specific model in tests." + }, + { + "left": "issue:44360", + "right": "issue:44561", + "accept": false, + "reason": "DSA/ReLU discussion is unrelated to the v5 trust_remote_code import breakage." + }, + { + "left": "issue:44857", + "right": "issue:45072", + "accept": false, + "reason": "Both involve precision issues, but they are model-specific crashes in different subsystems and not one fixable bug." + }, + { + "left": "issue:38175", + "right": "issue:44493", + "accept": false, + "reason": "Unexpected zero probabilities and unexpected position-id keys are different failure modes." + }, + { + "left": "issue:45920", + "right": "issue:46192", + "accept": false, + "reason": "Tokenizer ID regressions are not the same as return_tensors being ignored when text_kwargs is passed." + }, + { + "left": "issue:34689", + "right": "issue:35141", + "accept": false, + "reason": "Model-loading failure for Llama 3.2 Vision is unrelated to embedding resize reinitialization." + }, + { + "left": "issue:43701", + "right": "issue:45083", + "accept": false, + "reason": "Checkpoint key mismatch and qwen3_omni_moe feature-length behavior are different code-path problems." + }, + { + "left": "issue:44977", + "right": "issue:45229", + "accept": false, + "reason": "Flash-attention generation failure and multi-GPU OOM are unrelated bugs." + }, + { + "left": "issue:44448", + "right": "issue:44977", + "accept": false, + "reason": "Different model families and different regressions: generation/output mismatch vs flash-attention behavior." + }, + { + "left": "issue:44855", + "right": "issue:46399", + "accept": false, + "reason": "Both mention decorators/parsing, but one is a Python 3.13 script parse issue and the other is a missing registration decorator." + }, + { + "left": "issue:29127", + "right": "issue:32090", + "accept": false, + "reason": "LayoutLMv3 box error messaging is unrelated to Trainer GPU broadcast type errors." + }, + { + "left": "issue:43550", + "right": "issue:45083", + "accept": false, + "reason": "torch.compile/SDPA failure in Bamba is not the same as qwen3_omni_moe helper behavior." + }, + { + "left": "issue:43927", + "right": "issue:45698", + "accept": false, + "reason": "Saving custom token IDs and loading the wrong custom module are both save/load bugs, but not the same underlying issue." + }, + { + "left": "issue:42490", + "right": "issue:43575", + "accept": false, + "reason": "A trailing-slash import warning is unrelated to tensor-parallel OOM on Qwen2-57B." + }, + { + "left": "issue:40444", + "right": "issue:43526", + "accept": false, + "reason": "Multiple-image iterable dataset failures and BEiT reduce_labels behavior are different preprocessing bugs." + }, + { + "left": "issue:29942", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 test failures do not match the embedding reinit bug." + }, + { + "left": "issue:44561", + "right": "issue:44779", + "accept": false, + "reason": "A removed import helper breaking trust_remote_code is not the same as tokenizer output corruption." + }, + { + "left": "issue:27957", + "right": "issue:34689", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta is unrelated to Llama 3.2 Vision model loading." + }, + { + "left": "issue:44857", + "right": "issue:45588", + "accept": false, + "reason": "LwDetr AMP crashes and sink-less flash_attention AttributeError are different failures." + }, + { + "left": "issue:44977", + "right": "issue:45563", + "accept": false, + "reason": "Flash-attention generation correctness and a stale warning in paged generate are not the same bug." + }, + { + "left": "issue:44568", + "right": "issue:45491", + "accept": false, + "reason": "Tokenizer special-token handling and Gemma3 NaN embeddings have different causes and fixes." + }, + { + "left": "issue:28282", + "right": "issue:33357", + "accept": false, + "reason": "Missing PyTorch import and MacOS bus error on a CLIP model are unrelated load/runtime issues." + }, + { + "left": "issue:43278", + "right": "issue:43381", + "accept": false, + "reason": "BF16-to-FP32 dtype drift in eval is not the same as gradient checkpointing being disallowed in eval mode." + }, + { + "left": "issue:43412", + "right": "issue:45588", + "accept": false, + "reason": "Executorch export guards/erfinv support and flash_attention s_aux handling are different code paths." + }, + { + "left": "issue:45372", + "right": "issue:45588", + "accept": false, + "reason": "Missing ReasoningEffort import for Gemma 4 processors is unrelated to flash_attention sink-less crashes." + }, + { + "left": "issue:45588", + "right": "issue:45823", + "accept": false, + "reason": "Flash_attention AttributeError and Gemma4 PLE device mismatch are separate runtime bugs." + }, + { + "left": "issue:36010", + "right": "issue:37428", + "accept": false, + "reason": "Both are import errors, but they involve different removed symbols and different modules." + }, + { + "left": "issue:43746", + "right": "issue:46209", + "accept": false, + "reason": "PEFT local-checkpoint loading and FP8 inference on B200 are unrelated problems." + }, + { + "left": "issue:44448", + "right": "issue:44466", + "accept": false, + "reason": "Different output regression for Pegasus and lm_head serialization inconsistency are not the same defect." + }, + { + "left": "issue:45198", + "right": "issue:45803", + "accept": false, + "reason": "Wav2Vec2 save/tokenization failure is unrelated to Fuyu's bare except swallowing 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": 20, + "review_unit_index": 29, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:10000", + "issue:15354", + "issue:25251", + "issue:26413", + "issue:29127", + "issue:29942", + "issue:30064", + "issue:31501", + "issue:32090", + "issue:33357", + "issue:34567", + "issue:39692", + "issue:42490", + "issue:42915", + "issue:42994", + "issue:43065", + "issue:43066", + "issue:43278", + "issue:43381", + "issue:43575", + "issue:43611", + "issue:43653", + "issue:43701", + "issue:43873", + "issue:43976", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44265", + "issue:44361", + "issue:44442", + "issue:44462", + "issue:44479", + "issue:44734", + "issue:44743", + "issue:44871", + "issue:44977", + "issue:45198", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45412", + "issue:45440", + "issue:45459", + "issue:45563", + "issue:45712", + "issue:46121", + "issue:46369" + ], + "soft_pairs": [ + "issue:25251|issue:32090", + "issue:45245|issue:45563", + "issue:43653|issue:44743", + "issue:43065|issue:45412", + "issue:44186|issue:45198", + "issue:44361|issue:45198", + "issue:43611|issue:46121", + "issue:33357|issue:39692", + "issue:29127|issue:34567", + "issue:42915|issue:42994", + "issue:45459|issue:45712", + "issue:44871|issue:44977", + "issue:43976|issue:45440", + "issue:45412|issue:46369", + "issue:43278|issue:43575", + "issue:26413|issue:31501", + "issue:44206|issue:44479", + "issue:45372|issue:45459", + "issue:43994|issue:45412", + "issue:29942|issue:30064", + "issue:43066|issue:44462", + "issue:10000|issue:15354", + "issue:42490|issue:43873", + "issue:44265|issue:45198", + "issue:43381|issue:43701", + "issue:44734|issue:45406", + "issue:43653|issue:44442" + ], + "prepared_review_unit_hash": "582bf6d2056fe33afbc36b2e904511066d2f22cbd5117b4f2a8a5a178eba0a39", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35729, + "estimated_input_tokens": 8933, + "estimated_eval_tokens": 18122 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All suggested soft pairs are false matches: they describe different bugs or changes in different subsystems (tokenizers, training, serving, quantization, model loading, docs, and memory behavior). There is no coherent duplicate cluster or single canonical artifact here.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:25251", + "right": "issue:32090", + "accept": false, + "reason": "Different failures: pipeline output-shape regression vs Trainer broadcast TypeError." + }, + { + "left": "issue:45245", + "right": "issue:45563", + "accept": false, + "reason": "Unrelated: category-cardinality error vs stale warning in paged generate()." + }, + { + "left": "issue:43653", + "right": "issue:44743", + "accept": false, + "reason": "BigBird tokenizer special-token bug is unrelated to Qwen recurrent-state caching." + }, + { + "left": "issue:43065", + "right": "issue:45412", + "accept": false, + "reason": "Different code paths: Sam3PixelDecoder dummy conv vs RT-DETR GC/memory leak." + }, + { + "left": "issue:44186", + "right": "issue:45198", + "accept": false, + "reason": "LayoutLMv2 tokenizer NER/padding crash vs Wav2Vec2 save/tokenization failure." + }, + { + "left": "issue:44361", + "right": "issue:45198", + "accept": false, + "reason": "MLukeTokenizer AttributeError is unrelated to Wav2Vec2 save_pretrained/tokenization issues." + }, + { + "left": "issue:43611", + "right": "issue:46121", + "accept": false, + "reason": "Model loading with base_model_prefix is unrelated to rope-validation type handling." + }, + { + "left": "issue:33357", + "right": "issue:39692", + "accept": false, + "reason": "MacOS bus error on CLIP model differs from SigLIP2 doc/example errors." + }, + { + "left": "issue:29127", + "right": "issue:34567", + "accept": false, + "reason": "LayoutLMv3 box-clarity error is unrelated to TrainerState token-count tracking." + }, + { + "left": "issue:42915", + "right": "issue:42994", + "accept": false, + "reason": "FineGrainedFP8Config model failure is not the same as quantized-model save failure." + }, + { + "left": "issue:45459", + "right": "issue:45712", + "accept": false, + "reason": "Tokenizer exception masking vs leftover dummy classes leaking into repo checks." + }, + { + "left": "issue:44871", + "right": "issue:44977", + "accept": false, + "reason": "Gemma eos_token_id mismatch is unrelated to Qwen3.5 flash-attention generation." + }, + { + "left": "issue:43976", + "right": "issue:45440", + "accept": false, + "reason": "Python-version compatibility issue is not the same as DeepseekV3MoE divergence." + }, + { + "left": "issue:45412", + "right": "issue:46369", + "accept": false, + "reason": "RT-DETR object lifetime leak differs from sam3-litetext memory overflow." + }, + { + "left": "issue:43278", + "right": "issue:43575", + "accept": false, + "reason": "BF16/FP32 eval dtype change is unrelated to tp-related OOM on Qwen2-57B." + }, + { + "left": "issue:26413", + "right": "issue:31501", + "accept": false, + "reason": "Checkpoint resume iterator error is not the same as FineWeb training startup failure." + }, + { + "left": "issue:44206", + "right": "issue:44479", + "accept": false, + "reason": "Unsupported feature-extractor arg differs from video-input regression in Qwen variants." + }, + { + "left": "issue:45372", + "right": "issue:45459", + "accept": false, + "reason": "Gemma 4 processor import breakage is unrelated to protobuf error masking in tokenizers." + }, + { + "left": "issue:43994", + "right": "issue:45412", + "accept": false, + "reason": "SigLIP2 incorrect outputs are not a memory-release problem." + }, + { + "left": "issue:29942", + "right": "issue:30064", + "accept": false, + "reason": "Flash Attention 2 test failures are unrelated to void segmentation-map processing." + }, + { + "left": "issue:43066", + "right": "issue:44462", + "accept": false, + "reason": "Wrong tokenizer decoder type and ignoring tokenizer.json are different tokenizer-load bugs." + }, + { + "left": "issue:10000", + "right": "issue:15354", + "accept": false, + "reason": "DistilBert model issue in German is unrelated to torch.jit GeneratorExp export failure." + }, + { + "left": "issue:42490", + "right": "issue:43873", + "accept": false, + "reason": "Trailing-slash import warning is not the same as offloading/quantization behavior." + }, + { + "left": "issue:44265", + "right": "issue:45198", + "accept": false, + "reason": "torch.export + torch_compilable_check failure is unrelated to Wav2Vec2 save/tokenization failure." + }, + { + "left": "issue:43381", + "right": "issue:43701", + "accept": false, + "reason": "Gradient-checkpointing eval limitation is different from resume_from_checkpoint key mismatch." + }, + { + "left": "issue:44734", + "right": "issue:45406", + "accept": false, + "reason": "KV-cache continuation indexing crash is unrelated to Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:43653", + "right": "issue:44442", + "accept": false, + "reason": "BigBirdTokenizer mask-token registration bug is unrelated to AutoTokenizer loading FastSpeech2ConformerTokenizer." + } + ] + }, + "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": 20, + "review_unit_index": 30, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:29127", + "issue:29942", + "issue:32090", + "issue:33357", + "issue:35141", + "issue:41720", + "issue:42490", + "issue:42915", + "issue:43065", + "issue:43262", + "issue:43278", + "issue:43329", + "issue:43334", + "issue:43381", + "issue:43421", + "issue:43526", + "issue:43531", + "issue:43575", + "issue:43701", + "issue:43756", + "issue:43761", + "issue:43824", + "issue:43873", + "issue:43931", + "issue:43994", + "issue:44186", + "issue:44206", + "issue:44279", + "issue:44387", + "issue:44442", + "issue:44464", + "issue:44857", + "issue:44977", + "issue:45072", + "issue:45229", + "issue:45245", + "issue:45372", + "issue:45406", + "issue:45440", + "issue:45491", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45600", + "issue:45834", + "issue:45859", + "issue:46369" + ], + "soft_pairs": [ + "issue:43421|issue:45491", + "issue:43873|issue:45229", + "issue:43065|issue:43994", + "issue:27957|issue:29942", + "issue:32090|issue:33357", + "issue:43756|issue:45440", + "issue:45372|issue:45406", + "issue:43526|issue:45834", + "issue:41720|issue:45229", + "issue:43278|issue:43329", + "issue:45440|issue:45593", + "issue:43575|issue:43701", + "issue:43526|issue:44186", + "issue:42490|issue:43381", + "issue:43761|issue:45859", + "issue:43575|issue:44387", + "issue:43526|issue:44857", + "issue:27957|issue:33357", + "issue:43334|issue:43531", + "issue:44442|issue:45072", + "issue:44977|issue:45245", + "issue:43994|issue:46369", + "issue:43824|issue:44279", + "issue:42915|issue:43931", + "issue:27957|issue:29127", + "issue:29127|issue:35141", + "issue:43262|issue:44206", + "issue:44464|issue:45584", + "issue:45588|issue:45600" + ], + "prepared_review_unit_hash": "a007fa62dd03294648bc13a037dbcba6e3d686c49b94353e202f9d5890543103", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36353, + "estimated_input_tokens": 9089, + "estimated_eval_tokens": 18434 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is not a true duplicate set: the paired issues share only broad vocabulary (e.g. Flash Attention, loading, OOM, multimodal) but describe different models, code paths, and failure modes. No PRs are present; the broadest representative issue is the open quantization/offloading report.", + "confidence": 0.95, + "canonical_issue_reason": "issue:43873 is the broadest, still-open problem statement in the set and best represents the general runtime/memory-regression theme, though the cluster overall is heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43873 has the widest scope and highest likelihood of anchoring related follow-up work among these mostly unrelated issues.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43421", + "right": "issue:45491", + "accept": false, + "reason": "Different bugs: tokenizer post-processor update behavior vs Gemma3 sliding-window NaNs." + }, + { + "left": "issue:43873", + "right": "issue:45229", + "accept": false, + "reason": "Both mention memory pressure, but one is quantization offloading behavior and the other is multi-GPU OOM for Gemma4." + }, + { + "left": "issue:43065", + "right": "issue:43994", + "accept": false, + "reason": "Unrelated components: Sam3PixelDecoder dummy conv vs SigLIP2 wrong outputs with AutoModel/pipeline." + }, + { + "left": "issue:27957", + "right": "issue:29942", + "accept": false, + "reason": "Same Flash Attention 2 wording, but one is XLM-R support and the other is failing FA2 tests; different underlying issues." + }, + { + "left": "issue:32090", + "right": "issue:33357", + "accept": false, + "reason": "Different failures: Trainer NoneType broadcast error vs MacOS bus error with a community CLIP model." + }, + { + "left": "issue:43756", + "right": "issue:45440", + "accept": false, + "reason": "Different model families and bugs: Smollm3 RoPE-layer mismatch vs DeepseekV3 native implementation divergence." + }, + { + "left": "issue:45372", + "right": "issue:45406", + "accept": false, + "reason": "Gemma4 processor import dependency break vs Gemma4 processor missing _tokenizer; separate failures." + }, + { + "left": "issue:43526", + "right": "issue:45834", + "accept": false, + "reason": "Both are vision/OCR-ish, but one is BeitImageProcessorFast label reduction and the other is Kosmos2.5 long OCR index error." + }, + { + "left": "issue:41720", + "right": "issue:45229", + "accept": false, + "reason": "Different root causes: Qwen3 auto device mapping cudaErrorAssert vs Gemma4 multi-GPU OOM." + }, + { + "left": "issue:43278", + "right": "issue:43329", + "accept": false, + "reason": "Unrelated bugs: dtype drift in embedding layers vs undefined variables in multimodal token counting video path." + }, + { + "left": "issue:45440", + "right": "issue:45593", + "accept": false, + "reason": "DeepSeekV3 implementation mismatch is unrelated to D-FINE auxiliary-loss handling." + }, + { + "left": "issue:43575", + "right": "issue:43701", + "accept": false, + "reason": "Different problems: tensor-parallel OOM on Qwen2-57B vs resume_from_checkpoint key mismatch." + }, + { + "left": "issue:43526", + "right": "issue:44186", + "accept": false, + "reason": "Both involve processors/tokenizers, but BeitImageProcessorFast label reduction and LayoutLMv2 NER padding crashes are different bugs." + }, + { + "left": "issue:42490", + "right": "issue:43381", + "accept": false, + "reason": "Trailing-slash local model import warning is unrelated to gradient checkpointing in eval mode." + }, + { + "left": "issue:43761", + "right": "issue:45859", + "accept": false, + "reason": "Different code paths: CLIPVisionModel hidden_states regression vs Qwen3.5Moe CPU offload incompatibility." + }, + { + "left": "issue:43575", + "right": "issue:44387", + "accept": false, + "reason": "Both are memory-related, but one is model-loading OOM and the other is increased CUDA reserved memory under int4 quantization." + }, + { + "left": "issue:43526", + "right": "issue:44857", + "accept": false, + "reason": "Different image-processing bugs: label reduction vs image loss crashing under float16 AMP on CUDA." + }, + { + "left": "issue:27957", + "right": "issue:33357", + "accept": false, + "reason": "No shared underlying bug: Flash Attention 2 support for XLM-R vs MacOS bus error loading a CLIP model." + }, + { + "left": "issue:43334", + "right": "issue:43531", + "accept": false, + "reason": "Different Qwen3-VL vs Qwen3-MoE loading/runtime issues; one is missing pad_token_id, the other sliding_window behavior." + }, + { + "left": "issue:44442", + "right": "issue:45072", + "accept": false, + "reason": "Different tokenizer/model problems: FastSpeech2ConformerTokenizer loading vs dtype mismatch in SwitchTransformers/TimmWrapperModel." + }, + { + "left": "issue:44977", + "right": "issue:45245", + "accept": false, + "reason": "Qwen3.5 flash-attention generation bug is unrelated to category-count runtime limit." + }, + { + "left": "issue:43994", + "right": "issue:46369", + "accept": false, + "reason": "Different bugs: SigLIP2 output quality issue vs Sam3-lite text memory overflow." + }, + { + "left": "issue:43824", + "right": "issue:44279", + "accept": false, + "reason": "Importing Qwen2.5-VL is a specific API exposure issue; dependency issue is broader and not the same bug." + }, + { + "left": "issue:42915", + "right": "issue:43931", + "accept": false, + "reason": "Different Qwen3-Moe FP8 failure vs Qwen3-VL shape-mismatch loading error." + }, + { + "left": "issue:27957", + "right": "issue:29127", + "accept": false, + "reason": "XLM-R Flash Attention 2 support is unrelated to LayoutLMv3 box-validation error messaging." + }, + { + "left": "issue:29127", + "right": "issue:35141", + "accept": false, + "reason": "Different subsystems: LayoutLMv3 error clarity vs post_init reinitializing tied output embeddings." + }, + { + "left": "issue:43262", + "right": "issue:44206", + "accept": false, + "reason": "Audio chat-template sample-rate defaulting is unrelated to LasrFeatureExtractor center-arg crash." + }, + { + "left": "issue:44464", + "right": "issue:45584", + "accept": false, + "reason": "Chunked generation with compiled forward is a different failure mode from Whisper empty-transcription handling after token alignment." + }, + { + "left": "issue:45588", + "right": "issue:45600", + "accept": false, + "reason": "Flash-attention sink-less-model crash is unrelated to auto_mappings referencing removed Sam3LiteText configs." + } + ] + }, + "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": 20, + "review_unit_index": 31, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:28218", + "issue:32090", + "issue:33290", + "issue:34567", + "issue:34689", + "issue:36010", + "issue:42175", + "issue:42796", + "issue:42907", + "issue:43316", + "issue:43381", + "issue:43493", + "issue:43525", + "issue:43550", + "issue:43611", + "issue:43646", + "issue:43720", + "issue:43761", + "issue:43827", + "issue:43873", + "issue:43931", + "issue:44062", + "issue:44077", + "issue:44188", + "issue:44442", + "issue:44464", + "issue:44743", + "issue:44811", + "issue:44855", + "issue:44871", + "issue:44898", + "issue:44964", + "issue:44977", + "issue:45137", + "issue:45216", + "issue:45290", + "issue:45440", + "issue:45491", + "issue:45584", + "issue:45698", + "issue:45759", + "issue:45803", + "issue:45907", + "issue:46121", + "issue:46234", + "issue:46399" + ], + "soft_pairs": [ + "issue:44743|issue:44811", + "issue:27957|issue:33290", + "issue:43646|issue:44077", + "issue:44977|issue:45584", + "issue:28218|issue:32090", + "issue:33290|issue:34567", + "issue:44062|issue:45907", + "issue:44464|issue:45491", + "issue:44442|issue:44898", + "issue:43381|issue:43873", + "issue:46121|issue:46399", + "issue:43316|issue:43525", + "issue:44964|issue:45440", + "issue:34689|issue:43611", + "issue:32090|issue:33290", + "issue:43720|issue:44442", + "issue:43931|issue:45216", + "issue:34689|issue:36010", + "issue:43493|issue:43827", + "issue:43550|issue:44743", + "issue:42907|issue:45698", + "issue:44442|issue:45803", + "issue:42175|issue:46234", + "issue:42796|issue:45290", + "issue:43873|issue:44871", + "issue:43316|issue:45759", + "issue:43761|issue:44188", + "issue:44855|issue:45137" + ], + "prepared_review_unit_hash": "d7c9239a8bbbe2571c953df2aef4e152c7c4d6f9be76f53ae3bd5645c0ad8a97", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 28, + "serialized_chars": 35624, + "estimated_input_tokens": 8906, + "estimated_eval_tokens": 18068 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The items are a heterogeneous mix of unrelated issues; none of the proposed soft pairs look like the same underlying bug or change, so all soft edges should be rejected.", + "confidence": 0.98, + "canonical_issue_reason": "No single canonical issue stands out: this set spans unrelated tokenizer, training, loading, quantization, generation, and documentation bugs, so there is no cohesive duplicate cluster to collapse onto one issue.", + "canonical_pr_reason": "No PRs are present in the input.", + "best_issue_reason": "No best issue among these for deduplication purposes; the candidates are too diverse and do not share a common code path or fix.", + "best_pr_reason": "No PRs are present in the input.", + "soft_edge_verdicts": [ + { + "left": "issue:44743", + "right": "issue:44811", + "accept": false, + "reason": "Different components and failures: Qwen recurrent-state reset in cache handling vs Whisper batch decode token skipping." + }, + { + "left": "issue:27957", + "right": "issue:33290", + "accept": false, + "reason": "Flash Attention 2 model support and a Deepspeed/Adafactor OOM are unrelated bugs." + }, + { + "left": "issue:43646", + "right": "issue:44077", + "accept": false, + "reason": "Both concern initialization, but one is a broad Transformers 5 custom-model init regression and the other is a specific patchtsmixer post_init policy issue." + }, + { + "left": "issue:44977", + "right": "issue:45584", + "accept": false, + "reason": "Both are generation failures, but they affect different models and code paths: Qwen3.5 flash-attention vs Whisper empty-transcription handling." + }, + { + "left": "issue:28218", + "right": "issue:32090", + "accept": false, + "reason": "Tokenizer added-token spacing is unrelated to the Trainer/_gpu_broadcast_one NoneType error." + }, + { + "left": "issue:33290", + "right": "issue:34567", + "accept": false, + "reason": "An OOM with Adafactor in Deepspeed is not the same as TrainerState token counters not updating." + }, + { + "left": "issue:44062", + "right": "issue:45907", + "accept": false, + "reason": "AddedToken constructor argument collision and list_repo_files revision forwarding are unrelated API bugs." + }, + { + "left": "issue:44464", + "right": "issue:45491", + "accept": false, + "reason": "Compiled-forward chunked generation inconsistency and Gemma3 NaN embeddings from sliding-window attention are different failure modes." + }, + { + "left": "issue:44442", + "right": "issue:44898", + "accept": false, + "reason": "Tokenizer loading failure and Perceiver image classification resolution handling are unrelated subsystems." + }, + { + "left": "issue:43381", + "right": "issue:43873", + "accept": false, + "reason": "Gradient checkpointing in eval mode and quantization offloading are different behavioral issues." + }, + { + "left": "issue:46121", + "right": "issue:46399", + "accept": false, + "reason": "A JSON-list type error in rope validation is unrelated to Qwen3.5 rotary-embedding registration under use_kernels=True." + }, + { + "left": "issue:43316", + "right": "issue:43525", + "accept": false, + "reason": "Gemma3TextConfig API mismatch and Llama4Config missing pad_token_id are separate config bugs." + }, + { + "left": "issue:44964", + "right": "issue:45440", + "accept": false, + "reason": "Phi-4 multimodal model loading regression and DeepseekV3MoE divergence from remote implementation do not share the same fix." + }, + { + "left": "issue:34689", + "right": "issue:43611", + "accept": false, + "reason": "Both mention model loading breakage, but one is a version-specific Llama 3.2 vision regression and the other is a base_model_prefix-related Transformers 5 regression." + }, + { + "left": "issue:32090", + "right": "issue:33290", + "accept": false, + "reason": "Trainer broadcast type handling and Deepspeed Adafactor OOM are unrelated training/runtime bugs." + }, + { + "left": "issue:43720", + "right": "issue:44442", + "accept": false, + "reason": "BitNet packed-weight loading under accelerate and FastSpeech2ConformerTokenizer loading are entirely different code paths." + }, + { + "left": "issue:43931", + "right": "issue:45216", + "accept": false, + "reason": "Weight-shape mismatch when loading Qwen3-VL is not the same as save_pretrained producing an incorrect checkpoint." + }, + { + "left": "issue:34689", + "right": "issue:36010", + "accept": false, + "reason": "These are separate import/load regressions with different failure points and APIs." + }, + { + "left": "issue:43493", + "right": "issue:43827", + "accept": false, + "reason": "SigLIP2 implementation mismatch and docs still referencing pipeline() are unrelated." + }, + { + "left": "issue:43550", + "right": "issue:44743", + "accept": false, + "reason": "A torch.compile/SDPA failure in Bamba is unrelated to Qwen recurrent-state reset logic." + }, + { + "left": "issue:42907", + "right": "issue:45698", + "accept": false, + "reason": "Dequantized save failure and wrong custom module loading after save_pretrained are distinct serialization problems." + }, + { + "left": "issue:44442", + "right": "issue:45803", + "accept": false, + "reason": "Tokenizer loading and a bare-except swallowing errors in FuyuBatchFeature.convert_to_tensors() are not the same bug." + }, + { + "left": "issue:42175", + "right": "issue:46234", + "accept": false, + "reason": "Missing TensorFlow in a torch extra install and a missing import_utils helper are unrelated packaging/API issues." + }, + { + "left": "issue:42796", + "right": "issue:45290", + "accept": false, + "reason": "Ministral BPE marker handling and chat-template crashes on tool-call messages are different tokenizer behaviors." + }, + { + "left": "issue:43873", + "right": "issue:44871", + "accept": false, + "reason": "Quantization offloading and Gemma-3 eos_token_id inconsistency are separate configuration/runtime bugs." + }, + { + "left": "issue:43316", + "right": "issue:45759", + "accept": false, + "reason": "Gemma3TextConfig API discrepancy and Qwen multimodal text_config unwrapping are different config/model-construction issues." + }, + { + "left": "issue:43761", + "right": "issue:44188", + "accept": false, + "reason": "CLIPVisionModel hidden_states regression and attention-kernel divergence under torch.compile are unrelated." + }, + { + "left": "issue:44855", + "right": "issue:45137", + "accept": false, + "reason": "Python 3.13 torch.jit parsing/import error and DeepSpeed ZeRO3 deque underflow are different 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": 20, + "review_unit_index": 32, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:28282", + "issue:32090", + "issue:33290", + "issue:34689", + "issue:35141", + "issue:42175", + "issue:42617", + "issue:43295", + "issue:43316", + "issue:43540", + "issue:43550", + "issue:43575", + "issue:43606", + "issue:43611", + "issue:43643", + "issue:43650", + "issue:43701", + "issue:43716", + "issue:43742", + "issue:43825", + "issue:43931", + "issue:44077", + "issue:44162", + "issue:44190", + "issue:44464", + "issue:44479", + "issue:44496", + "issue:44514", + "issue:44568", + "issue:44610", + "issue:44792", + "issue:44871", + "issue:44913", + "issue:44987", + "issue:45479", + "issue:45584", + "issue:45632", + "issue:45803", + "issue:46178", + "issue:46202", + "issue:46234", + "issue:46291", + "issue:46365", + "issue:46369", + "issue:46421", + "issue:46424" + ], + "soft_pairs": [ + "issue:44190|issue:44496", + "issue:44464|issue:46421", + "issue:44987|issue:46178", + "issue:43643|issue:44913", + "issue:46234|issue:46291", + "issue:44077|issue:44479", + "issue:44568|issue:45479", + "issue:43295|issue:43825", + "issue:43611|issue:43931", + "issue:45632|issue:46424", + "issue:43550|issue:43701", + "issue:25251|issue:28282", + "issue:43606|issue:43701", + "issue:33290|issue:34689", + "issue:42617|issue:43575", + "issue:33290|issue:35141", + "issue:43575|issue:43742", + "issue:27957|issue:32090", + "issue:43316|issue:43716", + "issue:44610|issue:46369", + "issue:43650|issue:46202", + "issue:42175|issue:43316", + "issue:43295|issue:44871", + "issue:45803|issue:46365", + "issue:43540|issue:44514", + "issue:44871|issue:45584", + "issue:44162|issue:44792" + ], + "prepared_review_unit_hash": "4fd33aae28d7f1db2114a705e4aab628e5c856da9cf5ee3c2f8a97541efb77d1", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35741, + "estimated_input_tokens": 8936, + "estimated_eval_tokens": 18128 + }, + "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: model loading regressions, processor/tokenizer bugs, backend/import problems, and a few compile/cache/training edge cases. None of the soft pairs look like the same underlying bug or change, so I would reject all of them as duplicates.", + "confidence": 0.72, + "canonical_issue_reason": "If forced to pick one anchor, issue 43611 is the broadest user-facing regression in the set: a Transformers 5.0 model-loading break that could affect many models, unlike the more model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 43611 is the best representative issue for this cluster because it is a clear, high-impact regression and more general than the surrounding model-specific or narrow API problems.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44190", + "right": "issue:44496", + "accept": false, + "reason": "Local dataset loading failure and unrecognized-model/config loading failure are different bugs in different code paths." + }, + { + "left": "issue:44464", + "right": "issue:46421", + "accept": false, + "reason": "Both mention chunking/compiled execution, but one is about inconsistent outputs while the other is about cache lazy init causing recompilation." + }, + { + "left": "issue:44987", + "right": "issue:46178", + "accept": false, + "reason": "A model-loading regression for one repo and a backend-mapping error for tensorflow_text are unrelated failures." + }, + { + "left": "issue:43643", + "right": "issue:44913", + "accept": false, + "reason": "Both are config-related, but one is missing fields with trust_remote_code and the other is rotary_pct not persisting on reload." + }, + { + "left": "issue:46234", + "right": "issue:46291", + "accept": false, + "reason": "Missing import_utils helper and kernels version/revision import failure are different dependency/API issues." + }, + { + "left": "issue:44077", + "right": "issue:44479", + "accept": false, + "reason": "PatchTSMixer post_init gating and Qwen video-input regression are unrelated model-specific bugs." + }, + { + "left": "issue:44568", + "right": "issue:45479", + "accept": false, + "reason": "Tokenizer special-token behavior and sequence-classification zero-loss behavior are different concrete bugs." + }, + { + "left": "issue:43295", + "right": "issue:43825", + "accept": false, + "reason": "One is a processor/tokenizer API regression for custom model code; the other is an error-message issue in pipeline() about translation support." + }, + { + "left": "issue:43611", + "right": "issue:43931", + "accept": false, + "reason": "Both are loading failures, but one is a base_model_prefix regression and the other is a shape mismatch for a specific Qwen3-VL model." + }, + { + "left": "issue:45632", + "right": "issue:46424", + "accept": false, + "reason": "Cache-path collisions for trust_remote_code and max_cache_length being ignored except for quantized cache are different cache behaviors." + }, + { + "left": "issue:43550", + "right": "issue:43701", + "accept": false, + "reason": "torch.compile/SDPA failure and resume_from_checkpoint key mismatch are unrelated." + }, + { + "left": "issue:25251", + "right": "issue:28282", + "accept": false, + "reason": "Pipeline output nesting with top_k and missing PyTorch ImportError are not the same bug." + }, + { + "left": "issue:43606", + "right": "issue:43701", + "accept": false, + "reason": "CPU offload device mismatch and checkpoint key mismatch are unrelated training/runtime problems." + }, + { + "left": "issue:33290", + "right": "issue:34689", + "accept": false, + "reason": "AdaFactor OOM in DeepSpeed and Llama 3.2 Vision loading break are different regressions." + }, + { + "left": "issue:42617", + "right": "issue:43575", + "accept": false, + "reason": "A script execution problem and a tp-load OOM on a specific model are not the same issue." + }, + { + "left": "issue:33290", + "right": "issue:35141", + "accept": false, + "reason": "Optimizer memory blowup and embedding reinitialization on resize are unrelated bugs." + }, + { + "left": "issue:43575", + "right": "issue:43742", + "accept": false, + "reason": "Both are model-loading complaints, but one is OOM under tensor parallelism and the other is a key error for a different model." + }, + { + "left": "issue:27957", + "right": "issue:32090", + "accept": false, + "reason": "Flash Attention 2 support for XLM-Roberta and a Trainer broadcast TypeError are unrelated." + }, + { + "left": "issue:43316", + "right": "issue:43716", + "accept": false, + "reason": "Gemma3TextConfig API inconsistency and Mistral-3 image-preprocessor dtype mismatch are different subsystems." + }, + { + "left": "issue:44610", + "right": "issue:46369", + "accept": false, + "reason": "Processor output size mismatch and SAM memory overflow are not the same underlying defect." + }, + { + "left": "issue:43650", + "right": "issue:46202", + "accept": false, + "reason": "A placeholder/low-signal issue and a Gemma pipeline roundtrip bug do not describe the same problem." + }, + { + "left": "issue:42175", + "right": "issue:43316", + "accept": false, + "reason": "Packaging/backend omission and config API discrepancy are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:44871", + "accept": false, + "reason": "These both involve tokenizer/processor behavior, but one is a regression in processor.tokenizer/image passing and the other is an eos_token_id config inconsistency." + }, + { + "left": "issue:45803", + "right": "issue:46365", + "accept": false, + "reason": "Bare except swallowing interrupts and import failure until uninstalling datasets are different bugs." + }, + { + "left": "issue:43540", + "right": "issue:44514", + "accept": false, + "reason": "Qwen3OmniMoe video-processing error and batched apply_chat_template padding crash are separate code paths." + }, + { + "left": "issue:44871", + "right": "issue:45584", + "accept": false, + "reason": "Gemma eos_token_id inconsistency and Whisper empty-transcription failure after align_special_tokens are different model/tokenization bugs." + }, + { + "left": "issue:44162", + "right": "issue:44792", + "accept": false, + "reason": "ESM2 breakage and a janus image-generation test failure are unrelated model-specific 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": 20, + "review_unit_index": 33, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:41720", + "issue:41950", + "issue:43065", + "issue:43232", + "issue:43388", + "issue:43404", + "issue:43441", + "issue:43452", + "issue:43475", + "issue:43504", + "issue:43572", + "issue:43646", + "issue:43673", + "issue:43761", + "issue:43827", + "issue:43931", + "issue:43994", + "issue:44060", + "issue:44077", + "issue:44190", + "issue:44242", + "issue:44479", + "issue:44496", + "issue:44534", + "issue:44610", + "issue:44617", + "issue:44734", + "issue:44743", + "issue:44792", + "issue:44855", + "issue:44871", + "issue:44913", + "issue:44964", + "issue:44977", + "issue:45127", + "issue:45200", + "issue:45245", + "issue:45542", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45698", + "issue:45803", + "issue:46050", + "issue:46170", + "issue:46178", + "issue:46234", + "issue:46421" + ], + "soft_pairs": [ + "issue:45803|issue:46170", + "issue:41950|issue:43475", + "issue:44871|issue:46050", + "issue:43404|issue:43441", + "issue:43646|issue:44913", + "issue:43452|issue:44496", + "issue:43065|issue:44617", + "issue:45127|issue:45200", + "issue:43761|issue:44479", + "issue:43572|issue:46178", + "issue:43504|issue:45698", + "issue:45542|issue:46234", + "issue:43232|issue:43388", + "issue:44077|issue:44534", + "issue:44792|issue:44977", + "issue:43761|issue:44242", + "issue:44792|issue:45600", + "issue:43673|issue:46421", + "issue:44190|issue:44964", + "issue:44617|issue:45584", + "issue:43673|issue:44743", + "issue:44534|issue:44913", + "issue:44855|issue:45588", + "issue:44479|issue:44734", + "issue:43827|issue:43994", + "issue:43065|issue:44610", + "issue:41720|issue:45245", + "issue:43931|issue:44060" + ], + "prepared_review_unit_hash": "102552a2ef134deae8efa8d49fb6a7254ae5c03f826a4cfade35fcad80d6ddea", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36344, + "estimated_input_tokens": 9086, + "estimated_eval_tokens": 18428 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue set across generation, model loading, video, config, and backend regressions. The only strong duplicate-looking pair is the chunked-prefill cache bug (43673/46421).", + "confidence": 0.73, + "canonical_issue_reason": "issue:46421 is the best representative of the only clear duplicate pair: it uses the broader, more current wording around cache lazy initialization and chunked prefill.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:46421 is the most general and current issue among the likely duplicates, and it best captures the shared cache/chunked-prefill problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45803", + "right": "issue:46170", + "accept": false, + "reason": "Different bugs: exception handling in Fuyu preprocessing vs Windows/ROCm import dependency failure." + }, + { + "left": "issue:41950", + "right": "issue:43475", + "accept": false, + "reason": "Different pipelines/models: video-classification processor lookup vs SAM3 video missing attribute." + }, + { + "left": "issue:44871", + "right": "issue:46050", + "accept": false, + "reason": "Unrelated symptoms: Gemma eos_token_id mismatch vs quantized model save/load corruption." + }, + { + "left": "issue:43404", + "right": "issue:43441", + "accept": false, + "reason": "Different code paths: tied lm_head weights vs FlashAttention failure in Ministral-3." + }, + { + "left": "issue:43646", + "right": "issue:44913", + "accept": false, + "reason": "General v5 custom init breakage is not the same as GPTNeoX rotary_pct not persisting on reload." + }, + { + "left": "issue:43452", + "right": "issue:44496", + "accept": false, + "reason": "Different loading failures: gguf_file handling vs missing model_type in config.json." + }, + { + "left": "issue:43065", + "right": "issue:44617", + "accept": false, + "reason": "SAM3 PixelDecoder dummy layer issue is unrelated to SAM3Video OOM." + }, + { + "left": "issue:45127", + "right": "issue:45200", + "accept": false, + "reason": "LoRA/vocab tie-word-embeddings bug is unrelated to Gemma 4 mm_token_type_ids defaults." + }, + { + "left": "issue:43761", + "right": "issue:44479", + "accept": false, + "reason": "CLIP hidden_states regression and Qwen video input regression are different model behaviors." + }, + { + "left": "issue:43572", + "right": "issue:46178", + "accept": false, + "reason": "StableLm pad_token_idx config issue is unrelated to tensorflow_text backend mapping." + }, + { + "left": "issue:43504", + "right": "issue:45698", + "accept": false, + "reason": "Legacy field loading in BEiT is different from wrong custom module resolution after save/load." + }, + { + "left": "issue:45542", + "right": "issue:46234", + "accept": false, + "reason": "TensorBoard/TensorFlow backend detection bug is unrelated to missing import_utils helper." + }, + { + "left": "issue:43232", + "right": "issue:43388", + "accept": false, + "reason": "Generation kwargs syncing bug is unrelated to gather_for_metrics dropping tuple label elements." + }, + { + "left": "issue:44077", + "right": "issue:44534", + "accept": false, + "reason": "Optional post_init enforcement is unrelated to non-persistent buffer initialization corruption." + }, + { + "left": "issue:44792", + "right": "issue:44977", + "accept": false, + "reason": "Janus test failure and Qwen3.5 FlashAttention generation issue are different bugs." + }, + { + "left": "issue:43761", + "right": "issue:44242", + "accept": false, + "reason": "CLIP vision hidden_states regression is unrelated to router logits/load-balancing loss handling." + }, + { + "left": "issue:44792", + "right": "issue:45600", + "accept": false, + "reason": "A Janus model test failure is not the same as stale auto_mappings references in CI." + }, + { + "left": "issue:43673", + "right": "issue:46421", + "accept": true, + "reason": "Both describe the same chunked-prefill cache initialization problem in generation; one frames it as missing cache, the other as lazy init causing recompilation." + }, + { + "left": "issue:44190", + "right": "issue:44964", + "accept": false, + "reason": "Local dataset loading in a training script is unrelated to Phi-4 multimodal model loading." + }, + { + "left": "issue:44617", + "right": "issue:45584", + "accept": false, + "reason": "SAM3Video OOM is unrelated to Whisper empty transcription after align_special_tokens." + }, + { + "left": "issue:43673", + "right": "issue:44743", + "accept": false, + "reason": "Chunked-prefill cache initialization is not the same as recurrent states resetting with seq_len>1." + }, + { + "left": "issue:44534", + "right": "issue:44913", + "accept": false, + "reason": "Non-persistent buffer junk in v5 is unrelated to GPTNeoXConfig rotary_pct reload behavior." + }, + { + "left": "issue:44855", + "right": "issue:45588", + "accept": false, + "reason": "Python 3.13 torch.jit parsing error is unrelated to flash_attention s_aux=None crash." + }, + { + "left": "issue:44479", + "right": "issue:44734", + "accept": false, + "reason": "Video input regression in Qwen models is unrelated to KV-cache continuation tensor indexing in serve." + }, + { + "left": "issue:43827", + "right": "issue:43994", + "accept": false, + "reason": "Docs still referencing pipeline() is unrelated to SIGLIP2 producing nonsensical outputs." + }, + { + "left": "issue:43065", + "right": "issue:44610", + "accept": false, + "reason": "SAM3 PixelDecoder dummy layer issue is unrelated to OmDet-Turbo processor/model input-size mismatch." + }, + { + "left": "issue:41720", + "right": "issue:45245", + "accept": false, + "reason": "CUDA assert on Qwen3 auto device mapping is unrelated to category-count overflow." + }, + { + "left": "issue:43931", + "right": "issue:44060", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch is unrelated to Qwen3-Next tied-weights warning." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "Grounded and conservative overall. The summary matches the packet, and the only accepted soft pair (43673/46421) is reasonably supported by the titles as the same cache/chunked-prefill bug. The other rejections are appropriately cautious and not overstated." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 34, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:25251", + "issue:27957", + "issue:35141", + "issue:36246", + "issue:41628", + "issue:42490", + "issue:43116", + "issue:43295", + "issue:43335", + "issue:43404", + "issue:43540", + "issue:43550", + "issue:43644", + "issue:43976", + "issue:44077", + "issue:44112", + "issue:44190", + "issue:44246", + "issue:44462", + "issue:44521", + "issue:44617", + "issue:44734", + "issue:44737", + "issue:44779", + "issue:44855", + "issue:44877", + "issue:44913", + "issue:44938", + "issue:45003", + "issue:45127", + "issue:45216", + "issue:45245", + "issue:45276", + "issue:45357", + "issue:45381", + "issue:45542", + "issue:45584", + "issue:45600", + "issue:45632", + "issue:45803", + "issue:45907", + "issue:46082", + "issue:46121", + "issue:46234", + "issue:46369", + "issue:46421", + "issue:46489" + ], + "soft_pairs": [ + "issue:45381|issue:46369", + "issue:35141|issue:42490", + "issue:43976|issue:44855", + "issue:43644|issue:44913", + "issue:43550|issue:44190", + "issue:44077|issue:44877", + "issue:45127|issue:45357", + "issue:44617|issue:46369", + "issue:43540|issue:44734", + "issue:25251|issue:27957", + "issue:44779|issue:46489", + "issue:43335|issue:44190", + "issue:44112|issue:45600", + "issue:45632|issue:46234", + "issue:44462|issue:45907", + "issue:44737|issue:45803", + "issue:43116|issue:44190", + "issue:44938|issue:45542", + "issue:44521|issue:45381", + "issue:43295|issue:45276", + "issue:45276|issue:45584", + "issue:45632|issue:46421", + "issue:36246|issue:41628", + "issue:46082|issue:46121", + "issue:43404|issue:45216", + "issue:45245|issue:45584", + "issue:44246|issue:45003" + ], + "prepared_review_unit_hash": "c5203eebd6489b574618466d4f1e99c7e6775a8283ba812748266cfe28d5d01c", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 27, + "serialized_chars": 35423, + "estimated_input_tokens": 8856, + "estimated_eval_tokens": 17968 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Most pairs are only loosely related by subsystem or error type and should not be deduplicated. The only strong duplicate candidate is the DeepSeek tokenizer regression pair: issue 44779 and issue 46489 describe the same underlying v5 tokenizer breakage, with 46489 reading like a more specific follow-up to the broader 44779 report.", + "confidence": 0.87, + "canonical_issue_reason": "Issue 44779 is the best canonical because it is the broader, earlier DeepSeek tokenizer regression report; 46489 appears to be a narrower reproduction/follow-up for the same v5 incorrect-output bug.", + "canonical_pr_reason": null, + "best_issue_reason": "44779 is the strongest issue to keep as the umbrella duplicate target for the DeepSeek tokenizer regression cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45381", + "right": "issue:46369", + "accept": false, + "reason": "Both are multimodal/video-related, but one is a wrong vision-position-id bug and the other is a memory overflow/OOM report; not the same defect." + }, + { + "left": "issue:35141", + "right": "issue:42490", + "accept": false, + "reason": "These are unrelated: one is about output embeddings being reinitialized on resize, the other about a local-model path ending in a slash triggering an import warning." + }, + { + "left": "issue:43976", + "right": "issue:44855", + "accept": false, + "reason": "Both mention Python-version compatibility, but they are different failure modes in different code paths: package loading on Python 3.9+ vs an IndentationError in DebertaV2 import on Python 3.13." + }, + { + "left": "issue:43644", + "right": "issue:44913", + "accept": false, + "reason": "One is about non-persistent buffers being filled with junk in Transformers 5.0.0; the other is about a config value not persisting on reload. Similar theme, different bug." + }, + { + "left": "issue:43550", + "right": "issue:44190", + "accept": false, + "reason": "These are unrelated runtime/loading failures in different areas: Bamba SDPA torch.compile vs loading a local dataset in an example script." + }, + { + "left": "issue:44077", + "right": "issue:44877", + "accept": false, + "reason": "Both touch config validation, but the actual problems differ: optional post_init handling for patchtsmixer vs strict config blocking granite_speech loading." + }, + { + "left": "issue:45127", + "right": "issue:45357", + "accept": false, + "reason": "Both involve save/load or embedding-related regressions, but they affect different models and code paths; not mergeable as one concrete fix." + }, + { + "left": "issue:44617", + "right": "issue:46369", + "accept": false, + "reason": "Both are memory-related, but one is a specific CUDA OOM for Sam3Video and the other is a memory overflow in sam3-litetext; not the same bug." + }, + { + "left": "issue:43540", + "right": "issue:44734", + "accept": false, + "reason": "Both involve video/KV-cache handling at a high level, but they describe different crashes in different model/server paths." + }, + { + "left": "issue:25251", + "right": "issue:27957", + "accept": false, + "reason": "A pipeline output-shape regression and Flash Attention 2 support for XLMRoberta are unrelated." + }, + { + "left": "issue:44779", + "right": "issue:46489", + "accept": true, + "reason": "Both report DeepSeek tokenizer producing incorrect output in v5; 46489 explicitly frames the issue as a gap in the earlier fix, so they look like the same underlying regression." + }, + { + "left": "issue:43335", + "right": "issue:44190", + "accept": false, + "reason": "A SwitchTransformers sparse-layer construction bug and a local dataset loading issue are unrelated." + }, + { + "left": "issue:44112", + "right": "issue:45600", + "accept": false, + "reason": "Both mention CI, but one is a stale test failure in GraniteSpeech and the other is auto_mappings referencing removed configs; different artifacts and fixes." + }, + { + "left": "issue:45632", + "right": "issue:46234", + "accept": false, + "reason": "A cache-path collision for trust_remote_code and a missing import utility function are unrelated." + }, + { + "left": "issue:44462", + "right": "issue:45907", + "accept": false, + "reason": "Tokenizer repo resolution and list_repo_files revision forwarding are different hub/API issues." + }, + { + "left": "issue:44737", + "right": "issue:45803", + "accept": false, + "reason": "XLNet CPU allocation and a bare-except bug in Fuyu tensor conversion are unrelated." + }, + { + "left": "issue:43116", + "right": "issue:44190", + "accept": false, + "reason": "One is multi-label classification returning empty results, the other is a dataset-loading problem; not the same bug." + }, + { + "left": "issue:44938", + "right": "issue:45542", + "accept": false, + "reason": "Python 3.14 import/load failure and an undefined tf backend error from tensorboard-only installs are different compatibility issues." + }, + { + "left": "issue:44521", + "right": "issue:45381", + "accept": false, + "reason": "Both are multimodal-template related, but one is all-zero assistant masks and the other is wrong video position IDs; different outputs and code paths." + }, + { + "left": "issue:43295", + "right": "issue:45276", + "accept": false, + "reason": "A regression around processor.tokenizer/image passing and a gemma4 resize-token-embeddings bug are not the same defect." + }, + { + "left": "issue:45276", + "right": "issue:45584", + "accept": false, + "reason": "These are unrelated: embedding resize propagation vs Whisper generation failing on empty transcription." + }, + { + "left": "issue:45632", + "right": "issue:46421", + "accept": false, + "reason": "A trust_remote_code cache-path collision and spurious recompilation from cache lazy init are separate issues." + }, + { + "left": "issue:36246", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and different module surfaces." + }, + { + "left": "issue:46082", + "right": "issue:46121", + "accept": false, + "reason": "LlamaConfig head_dim validation and rope-parameter conversion on JSON-loaded lists are different configuration bugs." + }, + { + "left": "issue:43404", + "right": "issue:45216", + "accept": false, + "reason": "Both involve model save/tie-weight behavior, but one is missing lm_head tying in Mistral3 and the other is incorrect Qwen3.5 save_pretrained output; not the same bug." + }, + { + "left": "issue:45245", + "right": "issue:45584", + "accept": false, + "reason": "Category-count runtime limit and Whisper empty-transcription generation failure are unrelated." + }, + { + "left": "issue:44246", + "right": "issue:45003", + "accept": false, + "reason": "Slow import and unsafe sys.modules access are different import-time problems with different fixes." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is well-grounded in the packet and stays conservative on the soft-edge pairs. The only accepted duplicate candidate, issue 44779 vs 46489, is supported by the titles and the follow-up framing in 46489. No soft-edge acceptance appears overstated, and the non-duplicates are rejected for plausible, specific reasons." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 35, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:35141", + "issue:36246", + "issue:38617", + "issue:41950", + "issue:42175", + "issue:42796", + "issue:42907", + "issue:43295", + "issue:43299", + "issue:43381", + "issue:43388", + "issue:43441", + "issue:43479", + "issue:43526", + "issue:43531", + "issue:43643", + "issue:43650", + "issue:43673", + "issue:43688", + "issue:43854", + "issue:43901", + "issue:44016", + "issue:44112", + "issue:44186", + "issue:44220", + "issue:44279", + "issue:44336", + "issue:44464", + "issue:44496", + "issue:44521", + "issue:44683", + "issue:44734", + "issue:44792", + "issue:44933", + "issue:44964", + "issue:45042", + "issue:45216", + "issue:45335", + "issue:45362", + "issue:45375", + "issue:45446", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45593", + "issue:45600", + "issue:46424" + ], + "soft_pairs": [ + "issue:44933|issue:45362", + "issue:44186|issue:44734", + "issue:44683|issue:45446", + "issue:43673|issue:46424", + "issue:44464|issue:45563", + "issue:44964|issue:45600", + "issue:43854|issue:44964", + "issue:43643|issue:44496", + "issue:44112|issue:44964", + "issue:43650|issue:44220", + "issue:36246|issue:38617", + "issue:42796|issue:44521", + "issue:27957|issue:35141", + "issue:43479|issue:45375", + "issue:45563|issue:45584", + "issue:43531|issue:45600", + "issue:43295|issue:45335", + "issue:43526|issue:43901", + "issue:42907|issue:43441", + "issue:43441|issue:45588", + "issue:44496|issue:45042", + "issue:43673|issue:44336", + "issue:43299|issue:45216", + "issue:44792|issue:46424", + "issue:42175|issue:44279", + "issue:41950|issue:44016", + "issue:43381|issue:45593", + "issue:43901|issue:44496", + "issue:43388|issue:43688", + "issue:42796|issue:43650", + "issue:44016|issue:44220", + "issue:45335|issue:45584" + ], + "prepared_review_unit_hash": "96321b1e6e1e48dcec00089a6737592e6279d78b1e85d9523167f7af71309c16", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "serialized_chars": 37225, + "estimated_input_tokens": 9307, + "estimated_eval_tokens": 18870 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is a grab bag of unrelated issue reports, mostly around model loading, generation, configs, and a few packaging/import regressions. I do not see any soft pair that clearly describes the same underlying bug or change, so all candidate duplicate edges should be rejected.", + "confidence": 0.82, + "canonical_issue_reason": "44964 is the clearest, most general user-facing bug report in the set: a concrete model-loading regression against the latest transformers release. It is well-scoped and representative of the broader class of breakages here.", + "canonical_pr_reason": null, + "best_issue_reason": "44964 is the best standalone issue because it is specific, reproducible, and broadly relevant to users hitting model loading failures after upgrading.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44933", + "right": "issue:45362", + "accept": false, + "reason": "Different failures: missing image_utils import vs Qwen chat/runtime crash. No shared code path." + }, + { + "left": "issue:44186", + "right": "issue:44734", + "accept": false, + "reason": "Tokenizer padding/NER crash is unrelated to serve KV-cache continuation indexing." + }, + { + "left": "issue:44683", + "right": "issue:45446", + "accept": false, + "reason": "Both touch flex_attention, but one is a torch>=2.9 compile failure and the other is a version-check/import bug. Different concrete problems." + }, + { + "left": "issue:43673", + "right": "issue:46424", + "accept": false, + "reason": "Both mention cache, but one is missing GenerationMixin cache in chunked prefill and the other is max_cache_length only applying to quantized cache." + }, + { + "left": "issue:44464", + "right": "issue:45563", + "accept": false, + "reason": "Chunked generation inconsistency with compiled forward is unrelated to a stale warning about num_return_sequences." + }, + { + "left": "issue:44964", + "right": "issue:45600", + "accept": false, + "reason": "Different model/config regressions: Phi-4 multimodal loading vs removed Sam3LiteText auto-mapping references." + }, + { + "left": "issue:43854", + "right": "issue:44964", + "accept": false, + "reason": "Both are model-loading issues, but for different model families and likely different failures." + }, + { + "left": "issue:43643", + "right": "issue:44496", + "accept": false, + "reason": "AutoConfig trust_remote_code field loss is unrelated to unrecognized OLMo-Hybrid config loading." + }, + { + "left": "issue:44112", + "right": "issue:44964", + "accept": false, + "reason": "CI device override test failure is not the same as a model-load regression." + }, + { + "left": "issue:43650", + "right": "issue:44220", + "accept": false, + "reason": "No meaningful overlap: generic 'ADD THE DATA' issue vs fbank feature extraction bug." + }, + { + "left": "issue:36246", + "right": "issue:38617", + "accept": false, + "reason": "Separate import errors in different modules with no shared code path." + }, + { + "left": "issue:42796", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer BPE marker handling and multimodal assistant mask generation are different bugs." + }, + { + "left": "issue:27957", + "right": "issue:35141", + "accept": false, + "reason": "Flash Attention 2 support for XLM-R is unrelated to embedding reinitialization after resize_token_embeddings." + }, + { + "left": "issue:43479", + "right": "issue:45375", + "accept": false, + "reason": "Different config-field issues in different model families; not the same bug." + }, + { + "left": "issue:45563", + "right": "issue:45584", + "accept": false, + "reason": "A stale paging warning is unrelated to Whisper empty-transcription generation failure." + }, + { + "left": "issue:43531", + "right": "issue:45600", + "accept": false, + "reason": "Qwen3-MoE sliding_window behavior and Sam3LiteText auto-mapping cleanup are unrelated." + }, + { + "left": "issue:43295", + "right": "issue:45335", + "accept": false, + "reason": "Custom processor/tokenizer regression is unrelated to t5gemma decoder embedding resize behavior." + }, + { + "left": "issue:43526", + "right": "issue:43901", + "accept": false, + "reason": "BeitImageProcessorFast label reduction bug is unrelated to stale documentation wording." + }, + { + "left": "issue:42907", + "right": "issue:43441", + "accept": false, + "reason": "Both involve Ministral/Devstral or flash attention, but one is save/load of dequantized weights and the other is inference failure under FlashAttention; not the same bug." + }, + { + "left": "issue:43441", + "right": "issue:45588", + "accept": false, + "reason": "Both mention flash attention, but the reported failures are different: model-specific RC failure vs AttributeError on s_aux=None in sink-less models." + }, + { + "left": "issue:44496", + "right": "issue:45042", + "accept": false, + "reason": "Unrecognized model config and PIL image-processor torchvision dependency are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:44336", + "accept": false, + "reason": "Cache handling in generation is unrelated to ANSI escape codes in loading_report." + }, + { + "left": "issue:43299", + "right": "issue:45216", + "accept": false, + "reason": "Qwen3VL loading failure and Qwen3.5 save_pretrained checkpoint corruption are different issues." + }, + { + "left": "issue:44792", + "right": "issue:46424", + "accept": false, + "reason": "Janus image-generation test failure is unrelated to cache length configuration behavior." + }, + { + "left": "issue:42175", + "right": "issue:44279", + "accept": false, + "reason": "Package-extra backend composition issue and generic dependency issue are too broad and not the same concrete bug." + }, + { + "left": "issue:41950", + "right": "issue:44016", + "accept": false, + "reason": "Video-classification pipeline bug is unrelated to a notebook syntax error." + }, + { + "left": "issue:43381", + "right": "issue:45593", + "accept": false, + "reason": "Gradient checkpointing eval-mode limitation is unrelated to D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:43901", + "right": "issue:44496", + "accept": false, + "reason": "Documentation mismatch for return_all_scores is unrelated to an unrecognized model config error." + }, + { + "left": "issue:43388", + "right": "issue:43688", + "accept": false, + "reason": "Dropped labels in gather_for_metrics and auxiliary-loss normalization are different training bugs." + }, + { + "left": "issue:42796", + "right": "issue:43650", + "accept": false, + "reason": "Tokenizer BPE-marker handling has nothing to do with the generic 'ADD THE DATA' issue." + }, + { + "left": "issue:44016", + "right": "issue:44220", + "accept": false, + "reason": "Notebook syntax error and _torch_extract_fbank_features() are unrelated." + }, + { + "left": "issue:45335", + "right": "issue:45584", + "accept": false, + "reason": "t5gemma embedding resize behavior is unrelated to Whisper empty-transcription generation after align_special_tokens." + } + ] + }, + "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": 20, + "review_unit_index": 36, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:38175", + "issue:39692", + "issue:41950", + "issue:43295", + "issue:43299", + "issue:43441", + "issue:43450", + "issue:43479", + "issue:43525", + "issue:43577", + "issue:43650", + "issue:43824", + "issue:43854", + "issue:43901", + "issue:43976", + "issue:44016", + "issue:44062", + "issue:44075", + "issue:44242", + "issue:44279", + "issue:44297", + "issue:44351", + "issue:44368", + "issue:44462", + "issue:44466", + "issue:44496", + "issue:44617", + "issue:44683", + "issue:44779", + "issue:44933", + "issue:44987", + "issue:45020", + "issue:45103", + "issue:45200", + "issue:45216", + "issue:45276", + "issue:45278", + "issue:45357", + "issue:45372", + "issue:45405", + "issue:45542", + "issue:45759", + "issue:45812", + "issue:45823", + "issue:45834", + "issue:46082", + "issue:46209", + "issue:46399" + ], + "soft_pairs": [ + "issue:44496|issue:45759", + "issue:44075|issue:44368", + "issue:43577|issue:45216", + "issue:46082|issue:46399", + "issue:45103|issue:45372", + "issue:43824|issue:44933", + "issue:44279|issue:45405", + "issue:44242|issue:45200", + "issue:38175|issue:44617", + "issue:41950|issue:45020", + "issue:43299|issue:45357", + "issue:44351|issue:45278", + "issue:44016|issue:44062", + "issue:43650|issue:46209", + "issue:41950|issue:45823", + "issue:39692|issue:43901", + "issue:44683|issue:45542", + "issue:43295|issue:44779", + "issue:43976|issue:45405", + "issue:44779|issue:45812", + "issue:43976|issue:44279", + "issue:43299|issue:43854", + "issue:43479|issue:43525", + "issue:43976|issue:44987", + "issue:43450|issue:44279", + "issue:44297|issue:44462", + "issue:44466|issue:45276", + "issue:43441|issue:45834" + ], + "prepared_review_unit_hash": "1afc254e5954e0f6b24758bf9e97412de6826da9b1f0bdc61b9f9292eb144e90", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 36162, + "estimated_input_tokens": 9041, + "estimated_eval_tokens": 18338 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These are mostly independent Transformers issues spanning different models, configs, pipelines, and regression types. The soft-similar pairs share broad themes (loading, tokenizers, multimodal configs, version regressions) but not the same concrete bug, so none should be merged as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": "issue:45020 is the broadest umbrella regression report in the set and the best cluster anchor, though it does not actually duplicate the other issues.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45020 best represents the cluster\u2019s recurring theme of recent-version model-loading regressions.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44496", + "right": "issue:45759", + "accept": false, + "reason": "Both are model-loading/configuration failures, but one is an unrecognized model_type error and the other is a composite-config unwrap bug for Qwen multimodal configs." + }, + { + "left": "issue:44075", + "right": "issue:44368", + "accept": false, + "reason": "Unrelated problems: optimizer SGD arguments being ignored vs a tie_word_embeddings warning during Qwen fine-tuning." + }, + { + "left": "issue:43577", + "right": "issue:45216", + "accept": false, + "reason": "Different models and different bugs: BLIP2 dtype propagation on load vs incorrect Qwen3.5 checkpoint serialization." + }, + { + "left": "issue:46082", + "right": "issue:46399", + "accept": false, + "reason": "A config-validation issue for Llama vs a Qwen3.5 kernel registration crash; not the same code path." + }, + { + "left": "issue:45103", + "right": "issue:45372", + "accept": false, + "reason": "auto_docstring annotation handling crash vs a mistral_common import error affecting Gemma 4 processor loading." + }, + { + "left": "issue:43824", + "right": "issue:44933", + "accept": false, + "reason": "Both involve imports, but they are different missing symbols in different modules and not the same underlying export bug." + }, + { + "left": "issue:44279", + "right": "issue:45405", + "accept": false, + "reason": "A generic dependency issue is not the same as PEFT version pinning to an unreleased release." + }, + { + "left": "issue:44242", + "right": "issue:45200", + "accept": false, + "reason": "One is about router load-balancing loss, the other about defaulting mm_token_type_ids for text-only fine-tuning." + }, + { + "left": "issue:38175", + "right": "issue:44617", + "accept": false, + "reason": "Unrelated symptoms: zero probabilities in SigLIP2 vs CUDA OOM in Sam3Video." + }, + { + "left": "issue:41950", + "right": "issue:45020", + "accept": false, + "reason": "Video-classification looking for image processors is a pipeline routing bug, not the same as remote_code model loading regressions." + }, + { + "left": "issue:43299", + "right": "issue:45357", + "accept": false, + "reason": "Both concern Qwen multimodal models, but one is a loading failure and the other is incorrect save_pretrained key serialization." + }, + { + "left": "issue:44351", + "right": "issue:45278", + "accept": false, + "reason": "The specific HybridCache import error may be one symptom among many, but 45278 is a broad import-regression report rather than the same concrete bug." + }, + { + "left": "issue:44016", + "right": "issue:44062", + "accept": false, + "reason": "Notebook syntax error and AddedToken constructor argument collision are unrelated." + }, + { + "left": "issue:43650", + "right": "issue:46209", + "accept": false, + "reason": "Placeholder issue vs FP8 inference failure on B200; no shared bug." + }, + { + "left": "issue:41950", + "right": "issue:45823", + "accept": false, + "reason": "Video pipeline processor lookup and Gemma4 device placement during forward are distinct failures." + }, + { + "left": "issue:39692", + "right": "issue:43901", + "accept": false, + "reason": "Both are docs-related, but they describe different examples and different API mismatches." + }, + { + "left": "issue:44683", + "right": "issue:45542", + "accept": false, + "reason": "Compiled flex_attention on new torch versions is unrelated to a TensorFlow/backend dependency problem." + }, + { + "left": "issue:43295", + "right": "issue:44779", + "accept": false, + "reason": "Custom processor.tokenizer access regression and Deepseek tokenizer output regression are different code paths and different models." + }, + { + "left": "issue:43976", + "right": "issue:45405", + "accept": false, + "reason": "Python version support failure is not the same as a PEFT minimum-version bump." + }, + { + "left": "issue:44779", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but for different model families and different failure modes; not one duplicate bug." + }, + { + "left": "issue:43976", + "right": "issue:44279", + "accept": false, + "reason": "A runtime Python compatibility issue and a dependency issue are not the same underlying problem." + }, + { + "left": "issue:43299", + "right": "issue:43854", + "accept": false, + "reason": "Both are model-loading issues, but they affect different models and have different root causes." + }, + { + "left": "issue:43479", + "right": "issue:43525", + "accept": false, + "reason": "Default config initialization for Phi4Multimodal is unrelated to the Llama4Config missing pad_token_id attribute." + }, + { + "left": "issue:44466", + "right": "issue:45276", + "accept": false, + "reason": "Different bugs around embeddings/serialization: device-dependent lm_head.weight saving vs resize_token_embeddings not updating Gemma4 layer/output embeddings." + }, + { + "left": "issue:43441", + "right": "issue:45834", + "accept": false, + "reason": "FlashAttention failure in Ministral-3 and long-input OCR index error in Kosmos2.5 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": 20, + "review_unit_index": 37, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:33290", + "issue:41628", + "issue:42994", + "issue:43317", + "issue:43441", + "issue:43525", + "issue:43643", + "issue:43653", + "issue:43673", + "issue:43723", + "issue:43761", + "issue:43784", + "issue:43881", + "issue:43901", + "issue:43976", + "issue:43994", + "issue:44016", + "issue:44162", + "issue:44188", + "issue:44246", + "issue:44261", + "issue:44263", + "issue:44291", + "issue:44297", + "issue:44336", + "issue:44361", + "issue:44521", + "issue:44589", + "issue:44623", + "issue:44671", + "issue:44683", + "issue:44749", + "issue:44779", + "issue:45081", + "issue:45325", + "issue:45362", + "issue:45397", + "issue:45542", + "issue:45593", + "issue:45600", + "issue:45656", + "issue:45712", + "issue:45823", + "issue:45907", + "issue:46123", + "issue:46291", + "issue:46369", + "issue:46399" + ], + "soft_pairs": [ + "issue:45081|issue:45907", + "issue:43653|issue:45593", + "issue:44623|issue:44749", + "issue:42994|issue:43317", + "issue:45600|issue:45712", + "issue:44623|issue:45397", + "issue:44521|issue:45325", + "issue:33290|issue:45656", + "issue:44261|issue:44263", + "issue:46123|issue:46369", + "issue:44361|issue:45593", + "issue:44016|issue:45823", + "issue:43441|issue:44683", + "issue:44671|issue:44779", + "issue:43901|issue:43994", + "issue:43723|issue:44162", + "issue:46291|issue:46399", + "issue:44297|issue:45397", + "issue:43525|issue:43881", + "issue:43901|issue:44246", + "issue:44016|issue:44291", + "issue:43525|issue:43643", + "issue:44589|issue:45362", + "issue:44188|issue:44336", + "issue:41628|issue:43976", + "issue:43784|issue:45542", + "issue:43673|issue:43761" + ], + "prepared_review_unit_hash": "491902b07895feabb0afced74ca0e8093d3ff17ca9ead58db73d826a155af929", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 27, + "serialized_chars": 35786, + "estimated_input_tokens": 8947, + "estimated_eval_tokens": 18150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The set is heterogeneous, and every soft pair appears to be a false positive: they share broad themes like tokenizers, loading, or deepspeed, but not the same concrete bug/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:45081", + "right": "issue:45907", + "accept": false, + "reason": "Both touch tokenizer/revision handling, but one is a Mistral regex crash during load while the other is a repo_files API bug in CI." + }, + { + "left": "issue:43653", + "right": "issue:45593", + "accept": false, + "reason": "Completely different areas: BigBirdTokenizer special-token decode vs D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:44623", + "right": "issue:44749", + "accept": false, + "reason": "One is processor save_pretrained file generation; the other is a performance regression after upgrade." + }, + { + "left": "issue:42994", + "right": "issue:43317", + "accept": false, + "reason": "Both involve quantization/offload, but save failure and load failure are different code paths and symptoms." + }, + { + "left": "issue:45600", + "right": "issue:45712", + "accept": false, + "reason": "Both are CI/repo hygiene issues, but they involve different files and unrelated breakages." + }, + { + "left": "issue:44623", + "right": "issue:45397", + "accept": false, + "reason": "processor.save_pretrained missing files is unrelated to gemma-4 zero3 from_pretrained failures." + }, + { + "left": "issue:44521", + "right": "issue:45325", + "accept": false, + "reason": "Both are multimodal, but assistant_mask generation and rope index scaling are distinct bugs." + }, + { + "left": "issue:33290", + "right": "issue:45656", + "accept": false, + "reason": "Both mention deepspeed and optimizers, but OOM in Adafactor and double optimizer stepping are different problems." + }, + { + "left": "issue:44261", + "right": "issue:44263", + "accept": false, + "reason": "Different failures in different components: MLA layernorm precision vs torch.split return handling." + }, + { + "left": "issue:46123", + "right": "issue:46369", + "accept": false, + "reason": "MaskGenerationPipeline batch handling and sam3-litetext memory overflow are unrelated." + }, + { + "left": "issue:44361", + "right": "issue:45593", + "accept": false, + "reason": "MLukeTokenizer task errors do not match D-FINE auxiliary-loss behavior." + }, + { + "left": "issue:44016", + "right": "issue:45823", + "accept": false, + "reason": "A notebook syntax error and a Gemma4 device-mismatch forward bug are unrelated." + }, + { + "left": "issue:43441", + "right": "issue:44683", + "accept": false, + "reason": "Both concern attention kernels, but they affect different kernels/models and have different failure modes." + }, + { + "left": "issue:44671", + "right": "issue:44779", + "accept": false, + "reason": "Both are v5 regressions, but CamemBERT MLM output and Deepseek tokenizer correctness are different subsystems." + }, + { + "left": "issue:43901", + "right": "issue:43994", + "accept": false, + "reason": "Docs mismatch for return_all_scores is unrelated to incorrect model/pipeline outputs for SigLIP2." + }, + { + "left": "issue:43723", + "right": "issue:44162", + "accept": false, + "reason": "AutoTokenizer loading issues and ESM2 breakage are separate bugs." + }, + { + "left": "issue:46291", + "right": "issue:46399", + "accept": false, + "reason": "Both involve kernels, but one is a pipelines import dependency issue and the other is a Qwen3.5 registration crash." + }, + { + "left": "issue:44297", + "right": "issue:45397", + "accept": false, + "reason": "Tokenizer save metadata mismatch is unrelated to gemma-4 zero3 loading." + }, + { + "left": "issue:43525", + "right": "issue:43881", + "accept": false, + "reason": "Both are model-loading issues, but they involve different models and different missing-field/root causes." + }, + { + "left": "issue:43901", + "right": "issue:44246", + "accept": false, + "reason": "Pipeline docs wording and slow import are unrelated." + }, + { + "left": "issue:44016", + "right": "issue:44291", + "accept": false, + "reason": "Notebook syntax error and init_empty_weights argument handling are unrelated." + }, + { + "left": "issue:43525", + "right": "issue:43643", + "accept": false, + "reason": "Both mention missing config fields, but one is a specific model config attribute and the other is a trust_remote_code config-return bug." + }, + { + "left": "issue:44589", + "right": "issue:45362", + "accept": false, + "reason": "Storage-object import failure and Qwen3.5 chat crash are unrelated." + }, + { + "left": "issue:44188", + "right": "issue:44336", + "accept": false, + "reason": "Attention-kernel divergence and ANSI output formatting are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:43976", + "accept": false, + "reason": "Import error for AutoImageProcessor and Python-version compatibility are different issues." + }, + { + "left": "issue:43784", + "right": "issue:45542", + "accept": false, + "reason": "Both are dependency/import problems, but sentence-transformers nn import and TensorBoard TF backend handling are different code paths." + }, + { + "left": "issue:43673", + "right": "issue:43761", + "accept": false, + "reason": "Generation cache missing during chunked prefill is unrelated to CLIPVisionModel hidden_states 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": 20, + "review_unit_index": 38, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:16998", + "issue:22355", + "issue:36246", + "issue:41950", + "issue:42491", + "issue:42757", + "issue:42886", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43526", + "issue:43596", + "issue:43673", + "issue:43824", + "issue:43901", + "issue:43976", + "issue:44162", + "issue:44220", + "issue:44242", + "issue:44246", + "issue:44295", + "issue:44336", + "issue:44351", + "issue:44373", + "issue:44485", + "issue:44496", + "issue:44561", + "issue:44734", + "issue:44749", + "issue:44779", + "issue:44857", + "issue:44871", + "issue:44964", + "issue:45003", + "issue:45103", + "issue:45200", + "issue:45250", + "issue:45292", + "issue:45362", + "issue:45375", + "issue:45397", + "issue:45593", + "issue:45859", + "issue:46132", + "issue:46369", + "issue:46399", + "issue:46424" + ], + "soft_pairs": [ + "issue:44336|issue:44373", + "issue:16998|issue:22355", + "issue:45103|issue:46399", + "issue:44242|issue:45593", + "issue:44295|issue:45397", + "issue:44246|issue:44496", + "issue:42491|issue:43299", + "issue:44246|issue:44871", + "issue:44336|issue:44485", + "issue:43824|issue:43976", + "issue:41950|issue:44220", + "issue:44734|issue:46424", + "issue:36246|issue:42757", + "issue:36246|issue:44351", + "issue:44336|issue:46424", + "issue:43976|issue:44964", + "issue:44242|issue:44857", + "issue:42886|issue:46132", + "issue:44295|issue:44749", + "issue:43475|issue:46369", + "issue:45362|issue:45375", + "issue:45859|issue:46399", + "issue:44779|issue:45292", + "issue:43673|issue:44561", + "issue:43901|issue:45003", + "issue:43352|issue:45250", + "issue:44162|issue:45200", + "issue:36246|issue:43596", + "issue:43526|issue:45003", + "issue:44496|issue:45003" + ], + "prepared_review_unit_hash": "cce7cebeffe016de2a8dc96a3ea6fa15248be6b9b1c72202290b38150fdadb6b", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 30, + "serialized_chars": 35960, + "estimated_input_tokens": 8990, + "estimated_eval_tokens": 18236 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is heterogeneous: the soft links mostly connect unrelated import errors, model-specific regressions, docs/perf issues, and cache/config bugs. I found no true duplicate pair to merge, and there is no PR content here.", + "confidence": 0.95, + "canonical_issue_reason": "No clear canonical issue: the items span many unrelated bugs and regressions, so none is a safe duplicate representative.", + "canonical_pr_reason": null, + "best_issue_reason": "No single best issue candidate; this is not a coherent duplicate cluster, so choosing one issue would be misleading.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44336", + "right": "issue:44373", + "accept": false, + "reason": "ANSI-code rendering in loading_report is unrelated to a docstring typo for position_ids." + }, + { + "left": "issue:16998", + "right": "issue:22355", + "accept": false, + "reason": "model_max_length on DeBERTa-V3 is unrelated to the missing transformers.onnx module." + }, + { + "left": "issue:45103", + "right": "issue:46399", + "accept": false, + "reason": "auto_docstring annotation parsing and Qwen3.5 rotary-kernel registration are different code paths and bugs." + }, + { + "left": "issue:44242", + "right": "issue:45593", + "accept": false, + "reason": "Missing load-balancing loss in MoE routing is not the same as D-FINE auxiliary-loss handling when denoising is off." + }, + { + "left": "issue:44295", + "right": "issue:45397", + "accept": false, + "reason": "A position_ids buffer access error is unrelated to Gemma-4 zero3 from_pretrained loading." + }, + { + "left": "issue:44246", + "right": "issue:44496", + "accept": false, + "reason": "Import latency for transformers is unrelated to an unrecognized model/config error." + }, + { + "left": "issue:42491", + "right": "issue:43299", + "accept": false, + "reason": "Both mention Qwen3 MoE, but one is LoRA compatibility across versions and the other is model loading breakage; not the same concrete bug." + }, + { + "left": "issue:44246", + "right": "issue:44871", + "accept": false, + "reason": "Import performance and eos_token_id configuration mismatch are unrelated." + }, + { + "left": "issue:44336", + "right": "issue:44485", + "accept": false, + "reason": "Terminal ANSI output in loading_report has nothing to do with GLM-5 RoPE implementation." + }, + { + "left": "issue:43824", + "right": "issue:43976", + "accept": false, + "reason": "A missing Qwen2_5_VL export is unrelated to Python-version compatibility for Transformers 5.1.0." + }, + { + "left": "issue:41950", + "right": "issue:44220", + "accept": false, + "reason": "Video-classification image-processor lookup is unrelated to _torch_extract_fbank_features()." + }, + { + "left": "issue:44734", + "right": "issue:46424", + "accept": false, + "reason": "KV-cache continuation indexing in serve is different from max_cache_length semantics for Quantized Cache." + }, + { + "left": "issue:36246", + "right": "issue:42757", + "accept": false, + "reason": "A missing Qwen2_5_VL image processor export is unrelated to huggingface_hub's is_offline_mode import." + }, + { + "left": "issue:36246", + "right": "issue:44351", + "accept": false, + "reason": "Different missing symbols from different modules; not the same import bug." + }, + { + "left": "issue:43976", + "right": "issue:44964", + "accept": false, + "reason": "General Python-version support failure is unrelated to a specific Phi-4 multimodal loading error." + }, + { + "left": "issue:44242", + "right": "issue:44857", + "accept": false, + "reason": "MoE load-balancing loss logic and LWDETR AMP crash are different failures." + }, + { + "left": "issue:42886", + "right": "issue:46132", + "accept": false, + "reason": "Offline cache loading and AttentionInterface.register behavior are unrelated." + }, + { + "left": "issue:44295", + "right": "issue:44749", + "accept": false, + "reason": "A buffer-related position_ids error is not the same as slow filtering after upgrading Transformers." + }, + { + "left": "issue:43475", + "right": "issue:46369", + "accept": false, + "reason": "SAM3 encoder output missing an attribute is unrelated to memory overflow in sam3-litetext." + }, + { + "left": "issue:45362", + "right": "issue:45375", + "accept": false, + "reason": "A chat crash and a missing config field may be related at a high level, but they are not clearly the same concrete bug." + }, + { + "left": "issue:45859", + "right": "issue:46399", + "accept": false, + "reason": "CPU-offload incompatibility in text rotary embedding is different from the missing decorator error during apply_rotary_pos_emb registration." + }, + { + "left": "issue:44779", + "right": "issue:45292", + "accept": false, + "reason": "Deepseek tokenizer correctness and resize_token_embeddings output embedding syncing are different issues." + }, + { + "left": "issue:43673", + "right": "issue:44561", + "accept": false, + "reason": "Chunked-prefill cache absence is unrelated to removal of is_torch_fx_available breaking trust_remote_code models." + }, + { + "left": "issue:43901", + "right": "issue:45003", + "accept": false, + "reason": "A stale docstring about return_all_scores is unrelated to unsafe sys.modules access in modeling_utils." + }, + { + "left": "issue:43352", + "right": "issue:45250", + "accept": false, + "reason": "Flash-Attention support for Nemotron and a generic Flash Attention topic are not the same bug." + }, + { + "left": "issue:44162", + "right": "issue:45200", + "accept": false, + "reason": "ESM2 breakage and Gemma-4 mm_token_type_ids defaults are unrelated model-specific bugs." + }, + { + "left": "issue:36246", + "right": "issue:43596", + "accept": false, + "reason": "A Qwen2_5_VL import error is unrelated to a DeepSpeed zero3/BertModel IndexError." + }, + { + "left": "issue:43526", + "right": "issue:45003", + "accept": false, + "reason": "BeitImageProcessorFast label reduction and modeling_utils sys.modules access are different code paths." + }, + { + "left": "issue:44496", + "right": "issue:45003", + "accept": false, + "reason": "Unrecognized model/config loading and unsafe sys.modules access are unrelated 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": 20, + "review_unit_index": 39, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:43065", + "issue:43257", + "issue:43262", + "issue:43408", + "issue:43475", + "issue:43540", + "issue:43653", + "issue:43906", + "issue:43931", + "issue:44060", + "issue:44077", + "issue:44162", + "issue:44246", + "issue:44368", + "issue:44466", + "issue:44479", + "issue:44483", + "issue:44561", + "issue:44877", + "issue:44977", + "issue:44987", + "issue:44995", + "issue:45003", + "issue:45137", + "issue:45200", + "issue:45292", + "issue:45335", + "issue:45375", + "issue:45397", + "issue:45412", + "issue:45459", + "issue:45561", + "issue:45563", + "issue:45584", + "issue:45588", + "issue:45600", + "issue:45656", + "issue:45685", + "issue:45701", + "issue:45759", + "issue:45987", + "issue:46082", + "issue:46123", + "issue:46170", + "issue:46202", + "issue:46399", + "issue:46431" + ], + "soft_pairs": [ + "issue:43257|issue:45137", + "issue:45292|issue:45584", + "issue:45412|issue:46123", + "issue:43262|issue:44246", + "issue:44561|issue:46399", + "issue:45292|issue:45701", + "issue:45375|issue:45759", + "issue:44877|issue:45375", + "issue:45561|issue:45563", + "issue:44987|issue:45459", + "issue:44466|issue:45335", + "issue:44995|issue:45200", + "issue:45397|issue:45656", + "issue:45137|issue:45600", + "issue:45685|issue:46431", + "issue:43475|issue:43540", + "issue:44479|issue:46170", + "issue:43653|issue:44162", + "issue:44483|issue:46202", + "issue:44368|issue:44977", + "issue:44995|issue:45588", + "issue:44368|issue:45563", + "issue:43065|issue:43408", + "issue:43906|issue:46202", + "issue:45987|issue:46082", + "issue:43931|issue:44368", + "issue:44877|issue:46170", + "issue:44060|issue:44077", + "issue:45003|issue:45561" + ], + "prepared_review_unit_hash": "de55b7499700e20b9563dad6bd201fb3319dc385df64dcebcb0073a25b64f761", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35873, + "estimated_input_tokens": 8969, + "estimated_eval_tokens": 18194 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is highly heterogeneous: all soft links are broad topical similarities, not clear duplicates of the same bug or code change. I would reject every soft edge and avoid choosing a single canonical artifact for the set.", + "confidence": 0.94, + "canonical_issue_reason": "No clear canonical issue: the items cover unrelated failures across SAM3, Qwen, Gemma, DeepSpeed, tokenization, config loading, and test infrastructure.", + "canonical_pr_reason": null, + "best_issue_reason": "No single issue is a strong global representative because there is no coherent duplicate cluster; the titles point to distinct root causes and subsystems.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43257", + "right": "issue:45137", + "accept": false, + "reason": "Both mention DeepSpeed, but one is a Qwen3 MOE conversion bug and the other is a ZeRO3 deque crash; different code paths and symptoms." + }, + { + "left": "issue:45292", + "right": "issue:45584", + "accept": false, + "reason": "Resize-token-embeddings propagation and Whisper empty-transcription handling are unrelated bugs." + }, + { + "left": "issue:45412", + "right": "issue:46123", + "accept": false, + "reason": "Memory release after RT-DETR deletion is unrelated to MaskGenerationPipeline batch-finalization behavior." + }, + { + "left": "issue:43262", + "right": "issue:44246", + "accept": false, + "reason": "Audio sampling-rate defaults and slow import are different issues with no shared root cause." + }, + { + "left": "issue:44561", + "right": "issue:46399", + "accept": false, + "reason": "One is a v5 compatibility break in modeling_utils; the other is a Qwen3.5 kernel registration crash." + }, + { + "left": "issue:45292", + "right": "issue:45701", + "accept": false, + "reason": "Embedding resize propagation and tokenization changes across versions are not the same bug." + }, + { + "left": "issue:45375", + "right": "issue:45759", + "accept": false, + "reason": "Both involve Qwen multimodal config handling, but one is a missing config field under strict mode and the other is incorrect text_config unwrapping; not the same failure." + }, + { + "left": "issue:44877", + "right": "issue:45375", + "accept": false, + "reason": "Strict config loading breaks different model families for different reasons; these are not duplicates." + }, + { + "left": "issue:45561", + "right": "issue:45563", + "accept": false, + "reason": "A pytest-xdist file race and a stale generate() warning are unrelated." + }, + { + "left": "issue:44987", + "right": "issue:45459", + "accept": false, + "reason": "Model-load failure on a specific repo and protobuf-related tokenizer error masking are different underlying bugs." + }, + { + "left": "issue:44466", + "right": "issue:45335", + "accept": false, + "reason": "Both concern tied/embedding weights, but one is serialization inconsistency and the other is resize_token_embeddings not updating a decoder embedding." + }, + { + "left": "issue:44995", + "right": "issue:45200", + "accept": false, + "reason": "A stale cache crash in GlmMoeDsa is unrelated to missing mm_token_type_ids defaults." + }, + { + "left": "issue:45397", + "right": "issue:45656", + "accept": false, + "reason": "Gemma-4 ZeRO3 loading and duplicate optimizer stepping under deepspeed are different problems." + }, + { + "left": "issue:45137", + "right": "issue:45600", + "accept": false, + "reason": "A ZeRO3 deque error and stale auto_mappings references are separate CI/runtime issues." + }, + { + "left": "issue:45685", + "right": "issue:46431", + "accept": false, + "reason": "An MPS histogram integer-support error and missing chat template support are unrelated." + }, + { + "left": "issue:43475", + "right": "issue:43540", + "accept": false, + "reason": "Different SAM3/Qwen video-processing bugs with different object/model failures." + }, + { + "left": "issue:44479", + "right": "issue:46170", + "accept": false, + "reason": "A v5.3 video regression and a Windows ROCm import/dependency issue are unrelated." + }, + { + "left": "issue:43653", + "right": "issue:44162", + "accept": false, + "reason": "Tokenizer special-token registration and ESM2 breakage are not the same defect." + }, + { + "left": "issue:44483", + "right": "issue:46202", + "accept": false, + "reason": "An API request rejection and a Gemma roundtrip tokenization failure are different bugs." + }, + { + "left": "issue:44368", + "right": "issue:44977", + "accept": false, + "reason": "Both are Qwen3.5-related, but one is a warning about tied embeddings and the other is a flash-attention generation failure." + }, + { + "left": "issue:44995", + "right": "issue:45588", + "accept": false, + "reason": "GlmMoeDsa stale cache crash and flash_attention s_aux=None AttributeError are unrelated." + }, + { + "left": "issue:44368", + "right": "issue:45563", + "accept": false, + "reason": "A config warning and a paged-generation stale warning are not the same code-path failure." + }, + { + "left": "issue:43065", + "right": "issue:43408", + "accept": false, + "reason": "SAM3 pixel-decoder dummy conv and model-type mismatch warning are different SAM3 issues." + }, + { + "left": "issue:43906", + "right": "issue:46202", + "accept": false, + "reason": "Isolated reproduction of an older issue and a Gemma pipeline roundtrip bug are unrelated." + }, + { + "left": "issue:45987", + "right": "issue:46082", + "accept": false, + "reason": "StaticCache return-type contract and LlamaConfig head_dim validation are distinct model-internal issues." + }, + { + "left": "issue:43931", + "right": "issue:44368", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch and Qwen3.5 tied-weights warning do not share a concrete root cause." + }, + { + "left": "issue:44877", + "right": "issue:46170", + "accept": false, + "reason": "Strict config loading for granite_speech and Windows ROCm/torch.distributed import issues are unrelated." + }, + { + "left": "issue:44060", + "right": "issue:44077", + "accept": false, + "reason": "Qwen3-Next tied-weights warning and patchtsmixer post_init policy are not the same bug." + }, + { + "left": "issue:45003", + "right": "issue:45561", + "accept": false, + "reason": "Unsafe sys.modules access and an xdist file race are separate infrastructure 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": 20, + "review_unit_index": 40, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28282", + "issue:30990", + "issue:41628", + "issue:42371", + "issue:42673", + "issue:42886", + "issue:42994", + "issue:43232", + "issue:43475", + "issue:43482", + "issue:43525", + "issue:43643", + "issue:43646", + "issue:43673", + "issue:43701", + "issue:43704", + "issue:43784", + "issue:44038", + "issue:44162", + "issue:44188", + "issue:44297", + "issue:44351", + "issue:44361", + "issue:44623", + "issue:44749", + "issue:44869", + "issue:44871", + "issue:44933", + "issue:44964", + "issue:45081", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45229", + "issue:45230", + "issue:45310", + "issue:45356", + "issue:45375", + "issue:45381", + "issue:45397", + "issue:45478", + "issue:45584", + "issue:45656", + "issue:46132", + "issue:46170", + "issue:46202", + "issue:46399", + "issue:46421" + ], + "soft_pairs": [ + "issue:42886|issue:45229", + "issue:45584|issue:46170", + "issue:44297|issue:44623", + "issue:44361|issue:44749", + "issue:44749|issue:45656", + "issue:44623|issue:45656", + "issue:43701|issue:44749", + "issue:43646|issue:44162", + "issue:44869|issue:45356", + "issue:44871|issue:46132", + "issue:45230|issue:45397", + "issue:44188|issue:46202", + "issue:42994|issue:43482", + "issue:44038|issue:44623", + "issue:43475|issue:43643", + "issue:28282|issue:30990", + "issue:42371|issue:45092", + "issue:43475|issue:45381", + "issue:41628|issue:44933", + "issue:43232|issue:43673", + "issue:44188|issue:46421", + "issue:45375|issue:45478", + "issue:42994|issue:43525", + "issue:45310|issue:45375", + "issue:43784|issue:46399", + "issue:42673|issue:43704", + "issue:45081|issue:45103", + "issue:44964|issue:45137", + "issue:41628|issue:44351" + ], + "prepared_review_unit_hash": "c67c11a63efebc0efd1fceba07a5ff4de7a99379b66ce030fc035ded4c8a6b0d", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 29, + "serialized_chars": 36204, + "estimated_input_tokens": 9051, + "estimated_eval_tokens": 18358 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a heterogeneous cluster of unrelated Transformers issues, with only two clear duplicate subclusters standing out: the Qwen3.5 Moe from_pretrained regression and the repeated Qwen3ForCausalLM VRAM leak.", + "confidence": 0.58, + "canonical_issue_reason": "Issue 45375 is the best anchor here because it gives the concrete root cause for the Qwen3.5 Moe from_pretrained regression and best subsumes the generic 45310/45478 reports.", + "canonical_pr_reason": null, + "best_issue_reason": "45375 is the strongest issue to keep as canonical because it explains the regression instead of only restating the symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42886", + "right": "issue:45229", + "accept": false, + "reason": "Different bugs: offline cache tokenizer loading vs multi-GPU inference OOM." + }, + { + "left": "issue:45584", + "right": "issue:46170", + "accept": false, + "reason": "Unrelated failures: Whisper decoding crash vs Windows ROCm torch.distributed import error." + }, + { + "left": "issue:44297", + "right": "issue:44623", + "accept": false, + "reason": "Both involve save_pretrained, but one is tokenizer metadata and the other is processor file output." + }, + { + "left": "issue:44361", + "right": "issue:44749", + "accept": false, + "reason": "Different symptoms: tokenizer AttributeError vs a broad speed regression." + }, + { + "left": "issue:44749", + "right": "issue:45656", + "accept": false, + "reason": "Performance regression vs optimizer-step behavior; not the same code path." + }, + { + "left": "issue:44623", + "right": "issue:45656", + "accept": false, + "reason": "Processor save_pretrained bug vs optimizer double-step bug." + }, + { + "left": "issue:43701", + "right": "issue:44749", + "accept": false, + "reason": "Checkpoint resume key mismatch vs model loading slowdown." + }, + { + "left": "issue:43646", + "right": "issue:44162", + "accept": false, + "reason": "AutoConfig field-loss bug vs ESM2 model regression." + }, + { + "left": "issue:44869", + "right": "issue:45356", + "accept": false, + "reason": "Whisper timestamp decode crash vs Kimi tokenizer regression." + }, + { + "left": "issue:44871", + "right": "issue:46132", + "accept": false, + "reason": "Gemma eos_token_id mismatch vs AttentionInterface behavior change." + }, + { + "left": "issue:45230", + "right": "issue:45397", + "accept": false, + "reason": "Generic bug report vs specific Gemma-4 zero3 from_pretrained failure." + }, + { + "left": "issue:44188", + "right": "issue:46202", + "accept": false, + "reason": "Attention kernel divergence under torch.compile vs Gemma roundtrip token issue." + }, + { + "left": "issue:42994", + "right": "issue:43482", + "accept": false, + "reason": "Quantized model saving failure vs GGUF loading failure." + }, + { + "left": "issue:44038", + "right": "issue:44623", + "accept": false, + "reason": "Qwen3-VL-Moe model issue vs processor save_pretrained files." + }, + { + "left": "issue:43475", + "right": "issue:43643", + "accept": false, + "reason": "SAM3 encoder output attribute error vs AutoConfig field-loss bug." + }, + { + "left": "issue:28282", + "right": "issue:30990", + "accept": false, + "reason": "AutoModel missing PyTorch vs sentence-transformers loading stall." + }, + { + "left": "issue:42371", + "right": "issue:45092", + "accept": false, + "reason": "TF32 API migration issue vs InternVL2 meta-initialization incompatibility." + }, + { + "left": "issue:43475", + "right": "issue:45381", + "accept": false, + "reason": "SAM3 encoder output error vs qwen2.5-vl video position ids bug." + }, + { + "left": "issue:41628", + "right": "issue:44933", + "accept": false, + "reason": "Missing AutoImageProcessor import vs unrelated image_utils import issue." + }, + { + "left": "issue:43232", + "right": "issue:43673", + "accept": false, + "reason": "Generation kwarg update bug vs cache missing during chunked prefill; distinct paths." + }, + { + "left": "issue:44188", + "right": "issue:46421", + "accept": false, + "reason": "torch.compile attention branching issue vs cache lazy-init recompilation issue." + }, + { + "left": "issue:45375", + "right": "issue:45478", + "accept": true, + "reason": "Same Qwen3.5 Moe from_pretrained regression; 45375 is the root-cause report." + }, + { + "left": "issue:42994", + "right": "issue:43525", + "accept": false, + "reason": "Quantized save bug vs Llama4 config missing pad_token_id." + }, + { + "left": "issue:45310", + "right": "issue:45375", + "accept": true, + "reason": "Same bug family: generic Qwen3.5 Moe from_pretrained failure and the missing-field root cause." + }, + { + "left": "issue:43784", + "right": "issue:46399", + "accept": false, + "reason": "sentence-transformers import nn error vs Qwen3.5 kernel registration error." + }, + { + "left": "issue:42673", + "right": "issue:43704", + "accept": true, + "reason": "Exact duplicate Qwen3ForCausalLM VRAM leak report." + }, + { + "left": "issue:45081", + "right": "issue:45103", + "accept": false, + "reason": "Mistral regex patch crash vs auto_docstring future-annotations crash." + }, + { + "left": "issue:44964", + "right": "issue:45137", + "accept": false, + "reason": "Phi-4 multimodal load failure vs DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:41628", + "right": "issue:44351", + "accept": false, + "reason": "Missing AutoImageProcessor import vs HybridCache import error." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The analysis is well grounded in the packet and stays conservative on the soft edges. The accepted pairs are plausible duplicates/same-bug subclusters (45310/45375, 45375/45478, 42673/43704), and the rejects are generally well justified by distinct symptoms or code paths. The summary is appropriately cautious in describing the cluster as mostly heterogeneous with only those two clear duplicate subclusters." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 41, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:30333", + "issue:30990", + "issue:33290", + "issue:35141", + "issue:36010", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42831", + "issue:42898", + "issue:42994", + "issue:43065", + "issue:43232", + "issue:43278", + "issue:43377", + "issue:43723", + "issue:43784", + "issue:43824", + "issue:43874", + "issue:43901", + "issue:44016", + "issue:44038", + "issue:44188", + "issue:44222", + "issue:44242", + "issue:44373", + "issue:44496", + "issue:44568", + "issue:44623", + "issue:44734", + "issue:44779", + "issue:44792", + "issue:45092", + "issue:45103", + "issue:45137", + "issue:45200", + "issue:45310", + "issue:45356", + "issue:45406", + "issue:45561", + "issue:45600", + "issue:45712", + "issue:45812", + "issue:46077", + "issue:46123", + "issue:46202", + "issue:46231" + ], + "soft_pairs": [ + "issue:45406|issue:46202", + "issue:44222|issue:45310", + "issue:42994|issue:43824", + "issue:46077|issue:46123", + "issue:44038|issue:45310", + "issue:36010|issue:41628", + "issue:42831|issue:44222", + "issue:42175|issue:45712", + "issue:44496|issue:45561", + "issue:45561|issue:45600", + "issue:43901|issue:45561", + "issue:43232|issue:44792", + "issue:43377|issue:44242", + "issue:30990|issue:35141", + "issue:43278|issue:45137", + "issue:42898|issue:44568", + "issue:44623|issue:45310", + "issue:30333|issue:33290", + "issue:44188|issue:44373", + "issue:43065|issue:46123", + "issue:42222|issue:45200", + "issue:42898|issue:45812", + "issue:43723|issue:44568", + "issue:44016|issue:46231", + "issue:43784|issue:43824", + "issue:43874|issue:45092", + "issue:44734|issue:45103", + "issue:42898|issue:44779", + "issue:42898|issue:45356" + ], + "prepared_review_unit_hash": "2b131d7b373c37a31ed6e4ca9558186c7231d6bc6bdc72a670229f96693e0ded", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 29, + "serialized_chars": 35867, + "estimated_input_tokens": 8967, + "estimated_eval_tokens": 18190 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a set of superficially similar but substantively different bug reports; I would not merge any of the soft-edge pairs. The closest thematic overlap is among tokenizer/v5 regressions, but they still look like distinct code paths and fixes.", + "confidence": 0.79, + "canonical_issue_reason": "Issue #42898 is the clearest standalone regression report in the set and the most plausible hub for the tokenizer/v5 subset, but the overall cluster is too heterogeneous to treat as one duplicate group.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #42898 is the strongest representative issue: it is concrete, reproducible, and narrowly scoped to a specific tokenizer behavior regression.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45406", + "right": "issue:46202", + "accept": false, + "reason": "Different bugs: Gemma4Processor missing `_tokenizer` vs Gemma pipeline roundtrip on `\u2581`." + }, + { + "left": "issue:44222", + "right": "issue:45310", + "accept": false, + "reason": "Different code paths: FP8 `save_pretrained` in MoE vs `from_pretrained` failure for Qwen3.5 MoE." + }, + { + "left": "issue:42994", + "right": "issue:43824", + "accept": false, + "reason": "Unrelated failures: quantized model saving vs an import error for `Qwen2_5_VLForConditionalGeneration`." + }, + { + "left": "issue:46077", + "right": "issue:46123", + "accept": false, + "reason": "Different functionality: attention-mask construction vs batch-finalization in `MaskGenerationPipeline`." + }, + { + "left": "issue:44038", + "right": "issue:45310", + "accept": false, + "reason": "Both mention Qwen MoE, but the concrete failures differ and do not look like one fixable code-path bug." + }, + { + "left": "issue:36010", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but for different symbols (`GenerationMixin` vs `AutoImageProcessor`) and likely different causes." + }, + { + "left": "issue:42831", + "right": "issue:44222", + "accept": false, + "reason": "Accuracy regression in FineGrainedFP8 is not the same as a `save_pretrained` serialization failure." + }, + { + "left": "issue:42175", + "right": "issue:45712", + "accept": false, + "reason": "Packaging/backend dependency issue vs dummy-class cleanup/check-repo failure; not the same bug." + }, + { + "left": "issue:44496", + "right": "issue:45561", + "accept": false, + "reason": "Model config recognition failure and an xdist file race are unrelated." + }, + { + "left": "issue:45561", + "right": "issue:45600", + "accept": false, + "reason": "Test-worker race on `captured_info.txt` vs stale auto-mapping references; different root causes." + }, + { + "left": "issue:43901", + "right": "issue:45561", + "accept": false, + "reason": "Docs mismatch for pipeline return values vs a parallel test-file race." + }, + { + "left": "issue:43232", + "right": "issue:44792", + "accept": false, + "reason": "Generation kwargs update after `sync_gpus` is unrelated to the Janus image-generation test failure." + }, + { + "left": "issue:43377", + "right": "issue:44242", + "accept": false, + "reason": "Different model bugs: missing padding-mask support vs missing load-balancing loss when router logits are disabled." + }, + { + "left": "issue:30990", + "right": "issue:35141", + "accept": false, + "reason": "A loading hang and an embedding reinitialization bug are not the same underlying issue." + }, + { + "left": "issue:43278", + "right": "issue:45137", + "accept": false, + "reason": "Embedding dtype drift in eval is unrelated to the DeepSpeed ZeRO3 deque underflow." + }, + { + "left": "issue:42898", + "right": "issue:44568", + "accept": false, + "reason": "Both are tokenizer regressions, but one is cleanup behavior in v5 and the other is BOS/EOS insertion for a specific tokenizer." + }, + { + "left": "issue:44623", + "right": "issue:45310", + "accept": false, + "reason": "`processor.save_pretrained(...)` missing files is a save/export issue, not the same as a Qwen MoE load failure." + }, + { + "left": "issue:30333", + "right": "issue:33290", + "accept": false, + "reason": "MLflow reporting on failed training jobs is unrelated to an AdaFactor OOM in DeepSpeed." + }, + { + "left": "issue:44188", + "right": "issue:44373", + "accept": false, + "reason": "Attention-kernel divergence and a wrong docstring are completely different kinds of changes." + }, + { + "left": "issue:43065", + "right": "issue:46123", + "accept": false, + "reason": "Dummy `Conv2d` cleanup in Sam3PixelDecoder is unrelated to the partial-batch `is_last` bug." + }, + { + "left": "issue:42222", + "right": "issue:45200", + "accept": false, + "reason": "Broken VitPose models and Gemma4 `mm_token_type_ids` defaults are different model-specific problems." + }, + { + "left": "issue:42898", + "right": "issue:45812", + "accept": false, + "reason": "Both are tokenizer regressions, but they affect different models and token-ID logic, so they are not one duplicate." + }, + { + "left": "issue:43723", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer loading in v5 and missing BOS/EOS on `add_special_tokens=True` are different behaviors." + }, + { + "left": "issue:44016", + "right": "issue:46231", + "accept": false, + "reason": "Notebook syntax error and a docs parameter-name mismatch are unrelated documentation issues." + }, + { + "left": "issue:43784", + "right": "issue:43824", + "accept": false, + "reason": "Importing sentence-transformers with a missing `nn` symbol is unrelated to a missing Qwen VL export." + }, + { + "left": "issue:43874", + "right": "issue:45092", + "accept": false, + "reason": "Missing image-patch helper in GLM46V and InternVL2 meta-init incompatibility are different multimodal loading bugs." + }, + { + "left": "issue:44734", + "right": "issue:45103", + "accept": false, + "reason": "KV-cache continuation crash in serve is unrelated to the auto-docstring `__future__` annotations crash." + }, + { + "left": "issue:42898", + "right": "issue:44779", + "accept": false, + "reason": "Different tokenizer regressions affecting different models and code paths." + }, + { + "left": "issue:42898", + "right": "issue:45356", + "accept": false, + "reason": "Kimi-K2.5 codec handling and warning behavior is a different tokenizer regression than `clean_up_tokenization_spaces`." + } + ] + }, + "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": 20, + "review_unit_index": 42, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:30333", + "issue:34689", + "issue:35141", + "issue:36331", + "issue:41628", + "issue:42175", + "issue:42757", + "issue:42831", + "issue:42898", + "issue:43232", + "issue:43475", + "issue:43479", + "issue:43723", + "issue:43976", + "issue:44060", + "issue:44075", + "issue:44079", + "issue:44206", + "issue:44222", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44448", + "issue:44484", + "issue:44623", + "issue:44734", + "issue:44869", + "issue:44933", + "issue:44977", + "issue:45071", + "issue:45092", + "issue:45103", + "issue:45125", + "issue:45245", + "issue:45310", + "issue:45381", + "issue:45405", + "issue:45656", + "issue:45712", + "issue:46082", + "issue:46132", + "issue:46231", + "issue:46305", + "issue:46326", + "issue:46421", + "issue:46424" + ], + "soft_pairs": [ + "issue:44295|issue:45656", + "issue:43723|issue:46326", + "issue:42175|issue:44933", + "issue:42898|issue:44448", + "issue:44060|issue:45125", + "issue:45405|issue:45712", + "issue:44075|issue:46231", + "issue:44222|issue:44297", + "issue:43976|issue:44933", + "issue:44933|issue:45712", + "issue:44297|issue:45656", + "issue:27957|issue:30333", + "issue:44222|issue:44623", + "issue:30333|issue:34689", + "issue:44448|issue:46132", + "issue:42831|issue:44623", + "issue:43479|issue:44079", + "issue:30333|issue:36331", + "issue:30333|issue:35141", + "issue:45245|issue:46132", + "issue:43479|issue:44869", + "issue:44297|issue:45310", + "issue:41628|issue:42757", + "issue:45071|issue:46082", + "issue:43475|issue:46305", + "issue:43723|issue:44977", + "issue:44206|issue:45103", + "issue:43232|issue:46424", + "issue:45381|issue:46305", + "issue:44734|issue:46421", + "issue:44263|issue:45092", + "issue:44222|issue:44484" + ], + "prepared_review_unit_hash": "46b73aa037e7d70902bd388c8391c859dcd54e43e0eb09e2bd512d9362571596", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 32, + "serialized_chars": 36608, + "estimated_input_tokens": 9152, + "estimated_eval_tokens": 18560 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This is a loose set of Transformers issues with many lexical similarities but little evidence of true duplication. Most soft pairs are related only at a broad subsystem level (save_pretrained, cache, imports, generation, config) and should stay separate.", + "confidence": 0.77, + "canonical_issue_reason": "No strong canonical issue emerges: the items span unrelated bugs and regressions across model loading, tokenizers, training, cache, config, dependency, and docs paths, so there is no single clear duplicate anchor.", + "canonical_pr_reason": null, + "best_issue_reason": "If one issue must be used as a representative anchor, issue 44297 is the most concrete, user-facing serialization bug, but it still does not meaningfully subsume the rest of the set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44295", + "right": "issue:45656", + "accept": false, + "reason": "Different code paths: buffer registration/readout vs optimizer stepping under deepspeed. Same broad training/runtime area, not the same bug." + }, + { + "left": "issue:43723", + "right": "issue:46326", + "accept": false, + "reason": "Tokenizer loading regression vs mistral_common continue_final_message behavior. No shared underlying failure." + }, + { + "left": "issue:42175", + "right": "issue:44933", + "accept": false, + "reason": "Both are import/dependency complaints, but they concern different missing symbols/packages and unrelated fixes." + }, + { + "left": "issue:42898", + "right": "issue:44448", + "accept": false, + "reason": "One is a v5 tokenization cleanup behavior change; the other is a model output difference claim. Too indirect to treat as one bug." + }, + { + "left": "issue:44060", + "right": "issue:45125", + "accept": false, + "reason": "Different Qwen model bugs: a tied-weights warning issue versus missing tensor-parallel plan support." + }, + { + "left": "issue:45405", + "right": "issue:45712", + "accept": false, + "reason": "Unrelated release/versioning problem vs dummy-class cleanup failure." + }, + { + "left": "issue:44075", + "right": "issue:46231", + "accept": false, + "reason": "Optimizer behavior bug vs documentation parameter-name mismatch." + }, + { + "left": "issue:44222", + "right": "issue:44297", + "accept": false, + "reason": "Both mention save_pretrained, but one is FP8 MoE serialization and the other is tokenizer config metadata; not the same bug." + }, + { + "left": "issue:43976", + "right": "issue:44933", + "accept": false, + "reason": "Python version compatibility issue vs a missing image_utils import; unrelated causes and fixes." + }, + { + "left": "issue:44933", + "right": "issue:45712", + "accept": false, + "reason": "Different import-related problems: a missing image_utils export vs leftover dummy classes in pt objects." + }, + { + "left": "issue:44297", + "right": "issue:45656", + "accept": false, + "reason": "Tokenizer serialization metadata vs optimizer double-step under deepspeed are unrelated." + }, + { + "left": "issue:27957", + "right": "issue:30333", + "accept": false, + "reason": "Flash Attention model support vs MLflow reporting on failed jobs; no common underlying defect." + }, + { + "left": "issue:44222", + "right": "issue:44623", + "accept": false, + "reason": "Both involve save_pretrained, but one is FP8 MoE saving and the other is processor file emission; distinct bugs." + }, + { + "left": "issue:30333", + "right": "issue:34689", + "accept": false, + "reason": "Training-job/MLflow reporting bug vs Llama 3.2 Vision model loading regression." + }, + { + "left": "issue:44448", + "right": "issue:46132", + "accept": false, + "reason": "Version-difference output regression vs AttentionInterface.register behavior change; too broad and not the same code path." + }, + { + "left": "issue:42831", + "right": "issue:44623", + "accept": false, + "reason": "FP8 accuracy issue vs processor.save_pretrained file omission; unrelated symptoms." + }, + { + "left": "issue:43479", + "right": "issue:44079", + "accept": false, + "reason": "Multimodal config initialization bug vs ModelOutput key assignment bug." + }, + { + "left": "issue:30333", + "right": "issue:36331", + "accept": false, + "reason": "MLflow reporting failure vs Trainer compute_loss signature incompatibility." + }, + { + "left": "issue:30333", + "right": "issue:35141", + "accept": false, + "reason": "Failed-job reporting vs embedding reinitialization after resize; unrelated." + }, + { + "left": "issue:45245", + "right": "issue:46132", + "accept": false, + "reason": "Category cardinality limit error vs attention registration behavior change; no shared bug." + }, + { + "left": "issue:43479", + "right": "issue:44869", + "accept": false, + "reason": "Config defaulting bug vs Whisper timestamp decode crash; unrelated." + }, + { + "left": "issue:44297", + "right": "issue:45310", + "accept": false, + "reason": "Tokenizer save_pretrained metadata mismatch vs Qwen3.5 MoE from_pretrained failure." + }, + { + "left": "issue:41628", + "right": "issue:42757", + "accept": false, + "reason": "Different missing imports from different modules; same symptom class, not same defect." + }, + { + "left": "issue:45071", + "right": "issue:46082", + "accept": false, + "reason": "Type-checking regression vs LlamaConfig head_dim validation; unrelated." + }, + { + "left": "issue:43475", + "right": "issue:46305", + "accept": false, + "reason": "SAM 3 Video attribute error vs MoonshineStreaming mask off-by-one; different model families and code paths." + }, + { + "left": "issue:43723", + "right": "issue:44977", + "accept": false, + "reason": "Tokenizer loading regression vs Qwen3.5 flash-attention generation bug." + }, + { + "left": "issue:44206", + "right": "issue:45103", + "accept": false, + "reason": "Unsupported feature-extractor arg vs auto_docstring crash with future annotations; unrelated." + }, + { + "left": "issue:43232", + "right": "issue:46424", + "accept": false, + "reason": "Generation kwargs update bug vs cache max length constraint handling; both cache-related but not the same issue." + }, + { + "left": "issue:45381", + "right": "issue:46305", + "accept": false, + "reason": "Wrong vision_position_ids in Qwen2.5-VL video input vs MoonshineStreaming sliding-window off-by-one." + }, + { + "left": "issue:44734", + "right": "issue:46421", + "accept": false, + "reason": "Serving KV-cache indexing crash vs lazy cache initialization causing recompilation; related domain, different bugs." + }, + { + "left": "issue:44263", + "right": "issue:45092", + "accept": false, + "reason": "torch.split indexing logic bug vs remote-code checkpoint incompatibility with v5 meta init; unrelated." + }, + { + "left": "issue:44222", + "right": "issue:44484", + "accept": false, + "reason": "FP8 MoE save_pretrained bug vs max_shard_size default question; both touch saving, but not the same change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 43, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:22355", + "issue:27957", + "issue:28282", + "issue:30990", + "issue:38175", + "issue:39401", + "issue:41628", + "issue:42757", + "issue:42831", + "issue:42907", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43479", + "issue:43644", + "issue:43723", + "issue:43761", + "issue:43824", + "issue:43873", + "issue:44038", + "issue:44190", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44360", + "issue:44373", + "issue:44483", + "issue:44485", + "issue:44568", + "issue:44623", + "issue:44829", + "issue:44871", + "issue:44913", + "issue:44945", + "issue:45092", + "issue:45278", + "issue:45310", + "issue:45357", + "issue:45406", + "issue:45753", + "issue:45834", + "issue:46002", + "issue:46050", + "issue:46082", + "issue:46209", + "issue:46305", + "issue:46402", + "issue:46431" + ], + "soft_pairs": [ + "issue:43873|issue:44945", + "issue:46209|issue:46431", + "issue:44295|issue:44623", + "issue:44373|issue:44485", + "issue:44190|issue:46082", + "issue:43232|issue:45834", + "issue:43122|issue:46209", + "issue:43723|issue:44871", + "issue:44483|issue:45406", + "issue:44263|issue:44360", + "issue:43479|issue:43644", + "issue:42907|issue:46050", + "issue:30990|issue:38175", + "issue:44261|issue:45092", + "issue:45092|issue:45357", + "issue:43644|issue:46305", + "issue:42831|issue:45310", + "issue:44829|issue:46002", + "issue:22355|issue:28282", + "issue:44263|issue:44373", + "issue:42831|issue:44038", + "issue:41628|issue:43824", + "issue:43479|issue:44913", + "issue:43761|issue:46082", + "issue:43408|issue:46402", + "issue:22355|issue:27957", + "issue:44945|issue:45753", + "issue:39401|issue:44568", + "issue:42757|issue:45278", + "issue:27957|issue:30990" + ], + "prepared_review_unit_hash": "68265efbff3fe580f91868c266dd17b4de9eba88ae7c4276023b104f377d4bdf", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36321, + "estimated_input_tokens": 9081, + "estimated_eval_tokens": 18418 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is heterogeneous: most items are distinct regression reports, model-specific bugs, or unrelated import/config issues. None of the soft pairs look like the same underlying bug/change, so I rejected all soft edges. The broadest representative issue is the transformers v5 import-regression report.", + "confidence": 0.42, + "canonical_issue_reason": "issue:45278 is the broadest representative item in this set, covering upgrade-related import failures after moving to Transformers 5.5.0.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best cluster representative because it is broad, regression-oriented, and closer to a catch-all umbrella than the more model-specific reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43873", + "right": "issue:44945", + "accept": false, + "reason": "Different failures: quantization/offloading behavior versus pipeline-parallel output correctness." + }, + { + "left": "issue:46209", + "right": "issue:46431", + "accept": false, + "reason": "Unrelated subsystems: FP8 inference on B200 versus missing chat template in a processor." + }, + { + "left": "issue:44295", + "right": "issue:44623", + "accept": false, + "reason": "One is a buffer read/position_ids error; the other is processor.save_pretrained missing files." + }, + { + "left": "issue:44373", + "right": "issue:44485", + "accept": false, + "reason": "Docstring issue versus GLM-5 RoPE implementation bug; not the same code-path problem." + }, + { + "left": "issue:44190", + "right": "issue:46082", + "accept": false, + "reason": "Local dataset loading in an example script is unrelated to LlamaConfig head_dim validation." + }, + { + "left": "issue:43232", + "right": "issue:45834", + "accept": false, + "reason": "Generation kwargs/sync_gpus behavior is unrelated to Kosmos2.5 OCR index errors." + }, + { + "left": "issue:43122", + "right": "issue:46209", + "accept": false, + "reason": "Tokenizer version drift is unrelated to FP8 inference on Hopper/B200." + }, + { + "left": "issue:43723", + "right": "issue:44871", + "accept": false, + "reason": "Tokenizer loading regression and Gemma eos_token_id mismatch are different issues." + }, + { + "left": "issue:44483", + "right": "issue:45406", + "accept": false, + "reason": "Chat/completions request handling is not the same as Gemma4Processor missing _tokenizer." + }, + { + "left": "issue:44263", + "right": "issue:44360", + "accept": false, + "reason": "Both mention DSA/GlmMoe code, but they describe different defects and code paths." + }, + { + "left": "issue:43479", + "right": "issue:43644", + "accept": false, + "reason": "Default multimodal config initialization is unrelated to non-persistent buffer corruption on load." + }, + { + "left": "issue:42907", + "right": "issue:46050", + "accept": false, + "reason": "Different quantization/save-load failures on different models and mechanisms; not mergeable as one bug." + }, + { + "left": "issue:30990", + "right": "issue:38175", + "accept": false, + "reason": "Sentence Transformers loading hang is unrelated to zero probabilities in SIGLIP2 classification." + }, + { + "left": "issue:44261", + "right": "issue:45092", + "accept": false, + "reason": "Missing rms_norm_eps precision issue and meta-initialization checkpoint incompatibility are distinct bugs." + }, + { + "left": "issue:45092", + "right": "issue:45357", + "accept": false, + "reason": "Both are Qwen3.5-related, but one is checkpoint/meta-init incompatibility and the other is incorrect visual encoder keys on save." + }, + { + "left": "issue:43644", + "right": "issue:46305", + "accept": false, + "reason": "Non-persistent buffer initialization bug is unrelated to MoonshineStreaming\u2019s off-by-one mask window." + }, + { + "left": "issue:42831", + "right": "issue:45310", + "accept": false, + "reason": "FineGrainedFP8 accuracy problems are not the same as Qwen3.5 MoE from_pretrained errors." + }, + { + "left": "issue:44829", + "right": "issue:46002", + "accept": false, + "reason": "FlashAttention-3 degenerate training and NaNs after checkout/init are different training stability bugs." + }, + { + "left": "issue:22355", + "right": "issue:28282", + "accept": false, + "reason": "Missing transformers.onnx module is unrelated to AutoModel failing because PyTorch is absent." + }, + { + "left": "issue:44263", + "right": "issue:44373", + "accept": false, + "reason": "A tensor split bug and a docstring bug are not the same underlying issue." + }, + { + "left": "issue:42831", + "right": "issue:44038", + "accept": false, + "reason": "Different problems: FP8 accuracy versus a Qwen3-VL-Moe regression in Transformers 5." + }, + { + "left": "issue:41628", + "right": "issue:43824", + "accept": false, + "reason": "Import errors for different symbols are both broad import regressions, but they do not point to the same broken code path." + }, + { + "left": "issue:43479", + "right": "issue:44913", + "accept": false, + "reason": "Incorrect default multimodal config initialization is unrelated to GPTNeoX rotary_pct reload behavior." + }, + { + "left": "issue:43761", + "right": "issue:46082", + "accept": false, + "reason": "CLIP vision hidden_states regression and Llama head_dim validation are unrelated model-specific issues." + }, + { + "left": "issue:43408", + "right": "issue:46402", + "accept": false, + "reason": "Sam3 tracker/model-type warning is unrelated to text encoder weights not being loaded." + }, + { + "left": "issue:22355", + "right": "issue:27957", + "accept": false, + "reason": "Missing transformers.onnx and XLMRoberta Flash Attention 2 are distinct issues." + }, + { + "left": "issue:44945", + "right": "issue:45753", + "accept": false, + "reason": "Pipeline-parallel output correctness and Qwen3_5 image loop are different model/runtime bugs." + }, + { + "left": "issue:39401", + "right": "issue:44568", + "accept": false, + "reason": "Tokenizer offset_mapping bug is unrelated to add_special_tokens not adding BOS/EOS." + }, + { + "left": "issue:42757", + "right": "issue:45278", + "accept": false, + "reason": "A specific missing dependency symbol and a broad multi-import regression are related in theme only, not the same bug." + }, + { + "left": "issue:27957", + "right": "issue:30990", + "accept": false, + "reason": "Flash Attention 2 on XLMRoberta is unrelated to Sentence Transformers hanging during load." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 44, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:30990", + "issue:33357", + "issue:41628", + "issue:42371", + "issue:42796", + "issue:42886", + "issue:42907", + "issue:42994", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43232", + "issue:43408", + "issue:43452", + "issue:43475", + "issue:43550", + "issue:43576", + "issue:43646", + "issue:43723", + "issue:43756", + "issue:43784", + "issue:43937", + "issue:44038", + "issue:44060", + "issue:44117", + "issue:44261", + "issue:44263", + "issue:44295", + "issue:44297", + "issue:44360", + "issue:44368", + "issue:44373", + "issue:44485", + "issue:44514", + "issue:44625", + "issue:44704", + "issue:44829", + "issue:44945", + "issue:44964", + "issue:45200", + "issue:45278", + "issue:45290", + "issue:45593", + "issue:45600", + "issue:46002", + "issue:46132", + "issue:46305", + "issue:46369" + ], + "soft_pairs": [ + "issue:42886|issue:43066", + "issue:44368|issue:46305", + "issue:42796|issue:43723", + "issue:42886|issue:42994", + "issue:43408|issue:46369", + "issue:43550|issue:46132", + "issue:43646|issue:46002", + "issue:43576|issue:43784", + "issue:44295|issue:44297", + "issue:42371|issue:44263", + "issue:44117|issue:44704", + "issue:30990|issue:33357", + "issue:42907|issue:44945", + "issue:43232|issue:45593", + "issue:44261|issue:44373", + "issue:42796|issue:43122", + "issue:42907|issue:43937", + "issue:41628|issue:45278", + "issue:44704|issue:45290", + "issue:42907|issue:44964", + "issue:43784|issue:45600", + "issue:44038|issue:44297", + "issue:44360|issue:44485", + "issue:43116|issue:45200", + "issue:44038|issue:44295", + "issue:43756|issue:44060", + "issue:44625|issue:44829", + "issue:43408|issue:43475", + "issue:42907|issue:43452", + "issue:43122|issue:44514" + ], + "prepared_review_unit_hash": "3f4ccd3fc2c54a535ab646f50675c864ee6dcd98d56ecc5f88e4e8e8f36282e0", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36419, + "estimated_input_tokens": 9105, + "estimated_eval_tokens": 18466 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly unrelated issue reports, with a loose theme of Transformers v5 regressions around imports, tokenizers, and model loading. None of the soft-edge pairs look like true same-bug duplicates.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is the broadest umbrella report here: a general upgrade-to-v5 import regression that can reasonably represent the cluster better than the narrower, model-specific bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most representative user-facing failure in the set and the best anchor if one issue must stand in for the cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42886", + "right": "issue:43066", + "accept": false, + "reason": "Offline cache loading failure vs wrong decoder-type mapping; related tokenizer area, but different concrete bugs." + }, + { + "left": "issue:44368", + "right": "issue:46305", + "accept": false, + "reason": "Both mention warnings/behavior in model internals, but one is a tied-weights warning and the other is a sliding-window off-by-one." + }, + { + "left": "issue:42796", + "right": "issue:43723", + "accept": false, + "reason": "Specific tokenizer BPE-marker handling bug vs generic AutoTokenizer loading failure; not the same defect." + }, + { + "left": "issue:42886", + "right": "issue:42994", + "accept": false, + "reason": "Tokenizer offline-cache loading and quantized-model saving are unrelated failure modes." + }, + { + "left": "issue:43408", + "right": "issue:46369", + "accept": false, + "reason": "Same SAM3 family, but a type-mismatch warning is not the same as a memory overflow." + }, + { + "left": "issue:43550", + "right": "issue:46132", + "accept": false, + "reason": "One is a Bamba torch.compile/SDPA bug; the other is an AttentionInterface registration behavior change." + }, + { + "left": "issue:43646", + "right": "issue:46002", + "accept": false, + "reason": "Custom model initialization breakage and NaNs in classification heads are different code paths and symptoms." + }, + { + "left": "issue:43576", + "right": "issue:43784", + "accept": false, + "reason": "Broken env command vs import-time NameError in sentence-transformers; not the same bug." + }, + { + "left": "issue:44295", + "right": "issue:44297", + "accept": false, + "reason": "Position_ids buffer access error and tokenizer_class mismatch on save are separate tokenizer/model save issues." + }, + { + "left": "issue:42371", + "right": "issue:44263", + "accept": false, + "reason": "TF32 settings warning is unrelated to the DSA indexer torch.split issue." + }, + { + "left": "issue:44117", + "right": "issue:44704", + "accept": false, + "reason": "Mapping-lookup None handling and cached_file kwarg forwarding are different API bugs." + }, + { + "left": "issue:30990", + "right": "issue:33357", + "accept": false, + "reason": "SentenceTransformers loading hang and Mac bus error on CLIP are distinct runtime failures." + }, + { + "left": "issue:42907", + "right": "issue:44945", + "accept": false, + "reason": "Quantized-model save failure and pipeline-parallel output corruption are unrelated." + }, + { + "left": "issue:43232", + "right": "issue:45593", + "accept": false, + "reason": "Generation kwargs after sync_gpus and D-FINE auxiliary-loss omission do not share a code path." + }, + { + "left": "issue:44261", + "right": "issue:44373", + "accept": false, + "reason": "MLA precision/config issue and a wrong docstring are not duplicate bugs." + }, + { + "left": "issue:42796", + "right": "issue:43122", + "accept": false, + "reason": "Both involve tokenization, but one is a model-specific BPE marker bug and the other is a version-to-version tokenization change." + }, + { + "left": "issue:42907", + "right": "issue:43937", + "accept": false, + "reason": "Saving dequantized models and invalid GenerationConfig are unrelated." + }, + { + "left": "issue:41628", + "right": "issue:45278", + "accept": false, + "reason": "Missing AutoImageProcessor import is one instance of v5 import breakage, but 45278 is a broad umbrella rather than the same concrete bug." + }, + { + "left": "issue:44704", + "right": "issue:45290", + "accept": false, + "reason": "AutoProcessor kwarg forwarding and chat-template crashes with tool-call messages are different processor bugs." + }, + { + "left": "issue:42907", + "right": "issue:44964", + "accept": false, + "reason": "Quantized save failure and loading Phi-4 multimodal are unrelated model-loading problems." + }, + { + "left": "issue:43784", + "right": "issue:45600", + "accept": false, + "reason": "Sentence-transformers import NameError and stale auto_mappings references are different regression types." + }, + { + "left": "issue:44038", + "right": "issue:44297", + "accept": false, + "reason": "Qwen3-VL-Moe bug and tokenizer save metadata mismatch are not the same issue." + }, + { + "left": "issue:44360", + "right": "issue:44485", + "accept": false, + "reason": "DSA indexer activation omission and GLM-5 RoPE implementation are unrelated architecture bugs." + }, + { + "left": "issue:43116", + "right": "issue:45200", + "accept": false, + "reason": "Example-script multi-label output bug and Gemma 4 text-only token-type defaults are different behaviors." + }, + { + "left": "issue:44038", + "right": "issue:44295", + "accept": false, + "reason": "Qwen3-VL-Moe issue and position_ids buffer access error have no shared failure mechanism." + }, + { + "left": "issue:43756", + "right": "issue:44060", + "accept": false, + "reason": "Smollm3 RoPE layer mismatch and Qwen3-Next tied-weights warning are separate model-specific config problems." + }, + { + "left": "issue:44625", + "right": "issue:44829", + "accept": false, + "reason": "num_labels propagation and flash_attention_3 training collapse are unrelated bugs." + }, + { + "left": "issue:43408", + "right": "issue:43475", + "accept": false, + "reason": "Both are SAM 3 Video related, but model-type warning and missing fpn_position_embeddings are different defects." + }, + { + "left": "issue:42907", + "right": "issue:43452", + "accept": false, + "reason": "Quantized-model saving and gguf_file loading failures are distinct pre/post-save code paths." + }, + { + "left": "issue:43122", + "right": "issue:44514", + "accept": false, + "reason": "Tokenizer output differences across versions and batched chat-template crashes 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": 20, + "review_unit_index": 45, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:22355", + "issue:28218", + "issue:29127", + "issue:29159", + "issue:36296", + "issue:39401", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42757", + "issue:42831", + "issue:43066", + "issue:43116", + "issue:43122", + "issue:43441", + "issue:43493", + "issue:43611", + "issue:43824", + "issue:43825", + "issue:43874", + "issue:43937", + "issue:43976", + "issue:43992", + "issue:44038", + "issue:44075", + "issue:44222", + "issue:44279", + "issue:44297", + "issue:44351", + "issue:44355", + "issue:44485", + "issue:44488", + "issue:44521", + "issue:44704", + "issue:44779", + "issue:44936", + "issue:44938", + "issue:45083", + "issue:45216", + "issue:45357", + "issue:45446", + "issue:45478", + "issue:45563", + "issue:45753", + "issue:46002", + "issue:46231", + "issue:46328" + ], + "soft_pairs": [ + "issue:44038|issue:44222", + "issue:43122|issue:44779", + "issue:44279|issue:46328", + "issue:43122|issue:44521", + "issue:43824|issue:44351", + "issue:45216|issue:45478", + "issue:39401|issue:44488", + "issue:39401|issue:44521", + "issue:22355|issue:28218", + "issue:42757|issue:44351", + "issue:22355|issue:29159", + "issue:36296|issue:43116", + "issue:43992|issue:44704", + "issue:43066|issue:44779", + "issue:43937|issue:44485", + "issue:44355|issue:46231", + "issue:43611|issue:46002", + "issue:42371|issue:43874", + "issue:42831|issue:44297", + "issue:42222|issue:43441", + "issue:44038|issue:45478", + "issue:43976|issue:45446", + "issue:45357|issue:45478", + "issue:43493|issue:45753", + "issue:44075|issue:44936", + "issue:43116|issue:43825", + "issue:43116|issue:43493", + "issue:45083|issue:45446", + "issue:22355|issue:29127", + "issue:43937|issue:45563", + "issue:42175|issue:44938" + ], + "prepared_review_unit_hash": "7f9d60c97c7a6911934bb92b852cb77de5c4f2823479ffa6411b2d4984262046", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 31, + "serialized_chars": 35813, + "estimated_input_tokens": 8954, + "estimated_eval_tokens": 18164 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is not a true duplicate set; it mixes unrelated tokenizer regressions, model-specific load/save bugs, import errors, and documentation/performance issues. No soft edge looks like the same underlying bug/change with enough precision to merge.", + "confidence": 0.79, + "canonical_issue_reason": "issue:43122 is the broadest and most representative tokenizer-regression report in the set, so it is the least-bad anchor for the small tokenizer-related subgroup, though the overall cluster is still heterogeneous.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43122 is the best single representative issue because it has the clearest cross-version tokenizer regression framing and the strongest chance of being an umbrella for nearby tokenizer complaints; it still does not justify merging the whole cluster.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44038", + "right": "issue:44222", + "accept": false, + "reason": "Both mention moe models, but one is a Qwen3-VL-Moe load/inference bug and the other is FP8 save_pretrained behavior; different code paths and symptoms." + }, + { + "left": "issue:43122", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions in v5, but one is a generic cross-version tokenization change and the other is Deepseek-specific incorrect output; not the same concrete bug." + }, + { + "left": "issue:44279", + "right": "issue:46328", + "accept": false, + "reason": "A dependency complaint is unrelated to a SequenceFeatureExtractor.pad performance issue." + }, + { + "left": "issue:43122", + "right": "issue:44521", + "accept": false, + "reason": "Tokenization regression vs apply_chat_template assistant mask bug; different APIs and failure modes." + }, + { + "left": "issue:43824", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but for different symbols and modules; no shared underlying fix is evident." + }, + { + "left": "issue:45216", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen3.5-related, but one is save_pretrained checkpoint corruption and the other is from_pretrained loading failure; too different to merge." + }, + { + "left": "issue:39401", + "right": "issue:44488", + "accept": false, + "reason": "Qwen3 tokenizer offset_mapping bug vs loading failure for an unrelated model." + }, + { + "left": "issue:39401", + "right": "issue:44521", + "accept": false, + "reason": "Tokenizer offset mapping issue does not match a multimodal assistant_mask bug." + }, + { + "left": "issue:22355", + "right": "issue:28218", + "accept": false, + "reason": "ONNX import failure and added-token spacing are unrelated." + }, + { + "left": "issue:42757", + "right": "issue:44351", + "accept": false, + "reason": "Import errors, but from different packages and missing names; not the same defect." + }, + { + "left": "issue:22355", + "right": "issue:29159", + "accept": false, + "reason": "No module import problem vs slow/fast tokenizer inconsistency." + }, + { + "left": "issue:36296", + "right": "issue:43116", + "accept": false, + "reason": "Tensor parallel training bug and example-script classification output bug are unrelated." + }, + { + "left": "issue:43992", + "right": "issue:44704", + "accept": false, + "reason": "Both involve from_pretrained, but one misses a weight during model load while the other drops kwargs in processor file resolution." + }, + { + "left": "issue:43066", + "right": "issue:44779", + "accept": false, + "reason": "Wrong tokenizer decoder type serialization is different from Deepseek tokenization producing incorrect results." + }, + { + "left": "issue:43937", + "right": "issue:44485", + "accept": false, + "reason": "GenerationConfig validation error and GLM-5 RoPE implementation are distinct model issues." + }, + { + "left": "issue:44355", + "right": "issue:46231", + "accept": false, + "reason": "Running compiled Python files and a documentation parameter-name typo are unrelated." + }, + { + "left": "issue:43611", + "right": "issue:46002", + "accept": false, + "reason": "Base model prefix loading regression and NaNs in classification heads are different bugs." + }, + { + "left": "issue:42371", + "right": "issue:43874", + "accept": false, + "reason": "TF32 behavior docs/API issue vs missing image processor method; no overlap." + }, + { + "left": "issue:42831", + "right": "issue:44297", + "accept": false, + "reason": "FP8 accuracy issue and tokenizer_class mismatch are separate save/load concerns." + }, + { + "left": "issue:42222", + "right": "issue:43441", + "accept": false, + "reason": "VitPose model breakage and Ministral-3 FlashAttention failure are unrelated model-specific regressions." + }, + { + "left": "issue:44038", + "right": "issue:45478", + "accept": false, + "reason": "Both are Qwen/MoE related, but one is Qwen3-VL-Moe and the other is Qwen3.5 Moe from_pretrained; not the same concrete bug." + }, + { + "left": "issue:43976", + "right": "issue:45446", + "accept": false, + "reason": "Python version compatibility failure and a PyTorch version-check bug are unrelated." + }, + { + "left": "issue:45357", + "right": "issue:45478", + "accept": false, + "reason": "Both concern Qwen3.5 and version 5.x, but one is incorrect visual-encoder save_pretrained keys while the other is a from_pretrained load error." + }, + { + "left": "issue:43493", + "right": "issue:45753", + "accept": false, + "reason": "SigLIP2 implementation discrepancy and Qwen3_5 image infinite loop are unrelated model issues." + }, + { + "left": "issue:44075", + "right": "issue:44936", + "accept": false, + "reason": "Optimizer argument handling and trainer evaluate/train interaction are separate training bugs." + }, + { + "left": "issue:43116", + "right": "issue:43825", + "accept": false, + "reason": "Example multi-label classification returning empty results is unrelated to a pipeline error message about translation support." + }, + { + "left": "issue:43116", + "right": "issue:43493", + "accept": false, + "reason": "Classification example script bug and SigLIP2 implementation discrepancy are unrelated." + }, + { + "left": "issue:45083", + "right": "issue:45446", + "accept": false, + "reason": "qwen3_omni_moe feature-length helper behavior and flex_attention version gating are different code paths." + }, + { + "left": "issue:22355", + "right": "issue:29127", + "accept": false, + "reason": "Importing transformers.onnx and a LayoutLMv3 error-message clarity issue are unrelated." + }, + { + "left": "issue:43937", + "right": "issue:45563", + "accept": false, + "reason": "GLM generation config validation and a stale warning in paged generate() are different generation bugs." + }, + { + "left": "issue:42175", + "right": "issue:44938", + "accept": false, + "reason": "Backend packaging issue for pip install '.[torch]' and Python 3.14 load failure are unrelated environment 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": 20, + "review_unit_index": 46, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:33453", + "issue:38617", + "issue:39401", + "issue:41628", + "issue:42222", + "issue:42673", + "issue:43064", + "issue:43066", + "issue:43122", + "issue:43381", + "issue:43408", + "issue:43421", + "issue:43450", + "issue:43673", + "issue:43742", + "issue:43827", + "issue:43937", + "issue:43986", + "issue:44117", + "issue:44291", + "issue:44380", + "issue:44410", + "issue:44466", + "issue:44484", + "issue:44496", + "issue:44625", + "issue:44779", + "issue:44829", + "issue:44877", + "issue:44945", + "issue:44964", + "issue:44977", + "issue:45071", + "issue:45072", + "issue:45081", + "issue:45216", + "issue:45278", + "issue:45310", + "issue:45405", + "issue:45446", + "issue:45478", + "issue:45588", + "issue:46002", + "issue:46077", + "issue:46178", + "issue:46231", + "issue:46298", + "issue:46328" + ], + "soft_pairs": [ + "issue:38617|issue:45278", + "issue:33453|issue:39401", + "issue:44625|issue:46002", + "issue:43381|issue:44945", + "issue:44117|issue:45588", + "issue:44380|issue:46077", + "issue:39401|issue:44779", + "issue:43827|issue:45446", + "issue:43450|issue:43986", + "issue:43064|issue:43673", + "issue:44466|issue:46002", + "issue:43937|issue:44977", + "issue:45405|issue:45446", + "issue:44410|issue:45310", + "issue:43122|issue:46298", + "issue:38617|issue:41628", + "issue:44117|issue:45081", + "issue:42673|issue:45072", + "issue:42222|issue:46178", + "issue:44291|issue:45071", + "issue:44829|issue:45216", + "issue:43408|issue:43742", + "issue:44877|issue:46231", + "issue:44484|issue:45310", + "issue:39401|issue:43421", + "issue:43937|issue:44496", + "issue:44117|issue:46328", + "issue:43066|issue:43122", + "issue:44964|issue:45278", + "issue:44410|issue:45478" + ], + "prepared_review_unit_hash": "84ace9a682172bce6794b58ea46dd9d0c992a5919cbc31d42329591f243939cc", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 30, + "serialized_chars": 36702, + "estimated_input_tokens": 9176, + "estimated_eval_tokens": 18608 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster remains heterogeneous: it mixes distinct v5 regressions across tokenizers, model loading, attention/config handling, and docs. issue:45278 is still the best broad representative, but it is an umbrella rather than a true duplicate canonical.", + "confidence": 0.88, + "canonical_issue_reason": "issue:45278 is the broadest umbrella in the set, covering multiple import failures after upgrading to v5.5.0, so it best represents the cluster even though the items are not one duplicate bug.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most generally applicable issue to represent this mixed cluster, but it should be treated as a coarse umbrella, not a precise duplicate target.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38617", + "right": "issue:45278", + "accept": false, + "reason": "Both are import-related, but one is a specific missing symbol and the other is a broad upgrade regression; not the same concrete bug." + }, + { + "left": "issue:33453", + "right": "issue:39401", + "accept": false, + "reason": "Tokenizer loading regression vs. Qwen3 offset_mapping bug are different symptoms and likely different code paths." + }, + { + "left": "issue:44625", + "right": "issue:46002", + "accept": false, + "reason": "num_labels propagation and NaNs in classification heads are unrelated model-behavior issues." + }, + { + "left": "issue:43381", + "right": "issue:44945", + "accept": false, + "reason": "Eval-mode gradient checkpointing failure and pipeline-parallel wrong outputs are distinct training/runtime problems." + }, + { + "left": "issue:44117", + "right": "issue:45588", + "accept": false, + "reason": "Tokenizer mapping fallback and flash-attention crash on s_aux=None affect different subsystems." + }, + { + "left": "issue:44380", + "right": "issue:46077", + "accept": false, + "reason": "Both concern attention behavior, but they involve different models and different mechanisms." + }, + { + "left": "issue:39401", + "right": "issue:44779", + "accept": false, + "reason": "Both are tokenizer regressions, but they are model-specific incorrect-tokenization reports without evidence of a shared root cause." + }, + { + "left": "issue:43827", + "right": "issue:45446", + "accept": false, + "reason": "Docs still referencing pipeline() and a PyTorch version-check bug are unrelated." + }, + { + "left": "issue:43450", + "right": "issue:43986", + "accept": false, + "reason": "Batched video shape bug and missing-torchvision crash are different video-processing failures." + }, + { + "left": "issue:43064", + "right": "issue:43673", + "accept": false, + "reason": "Optimizer-state corruption in training and missing generation cache during chunked prefill are unrelated." + }, + { + "left": "issue:44466", + "right": "issue:46002", + "accept": false, + "reason": "Serialization of tied weights and NaNs in classification heads are different failure modes." + }, + { + "left": "issue:43937", + "right": "issue:44977", + "accept": false, + "reason": "Invalid GenerationConfig for GLM-5 and Qwen3.5 flash-attention generation failure are different issues." + }, + { + "left": "issue:45405", + "right": "issue:45446", + "accept": false, + "reason": "A released-version dependency bump and a PyTorch version check bug are not the same bug." + }, + { + "left": "issue:44410", + "right": "issue:45310", + "accept": false, + "reason": "Missing qkv/gate projections in qwen3next and a Qwen3.5 MoE from_pretrained error are different model-loading defects." + }, + { + "left": "issue:43122", + "right": "issue:46298", + "accept": false, + "reason": "Tokenization regression and noisy logging are unrelated." + }, + { + "left": "issue:38617", + "right": "issue:41628", + "accept": false, + "reason": "Both are import errors, but they involve different missing symbols and likely different regression causes." + }, + { + "left": "issue:44117", + "right": "issue:45081", + "accept": false, + "reason": "Tokenizer mapping None-handling and Mistral regex patch crash are different tokenizer-loading bugs." + }, + { + "left": "issue:42673", + "right": "issue:45072", + "accept": false, + "reason": "VRAM leak in Qwen3ForCausalLM threading and bfloat16 dtype mismatch in CI are unrelated." + }, + { + "left": "issue:42222", + "right": "issue:46178", + "accept": false, + "reason": "Broken vitpose models and an unsupported backend error are unrelated." + }, + { + "left": "issue:44291", + "right": "issue:45071", + "accept": false, + "reason": "Unexpected init_empty_weights argument and PretrainedConfig type-checking breakage are both init-path regressions, but not the same concrete bug." + }, + { + "left": "issue:44829", + "right": "issue:45216", + "accept": false, + "reason": "FlashAttention-3 degenerate training and Qwen3.5 save_pretrained regression are unrelated." + }, + { + "left": "issue:43408", + "right": "issue:43742", + "accept": false, + "reason": "Wrong model-type warning and MobileLLM key error are different model-loading issues." + }, + { + "left": "issue:44877", + "right": "issue:46231", + "accept": false, + "reason": "Strict config loading and a documentation parameter-name typo are unrelated." + }, + { + "left": "issue:44484", + "right": "issue:45310", + "accept": false, + "reason": "save_pretrained shard-size behavior and Qwen3.5 from_pretrained failure are different bugs." + }, + { + "left": "issue:39401", + "right": "issue:43421", + "accept": false, + "reason": "Qwen3 offset mapping and runtime special-token post-processor updates are different tokenizer behaviors." + }, + { + "left": "issue:43937", + "right": "issue:44496", + "accept": false, + "reason": "Invalid generation config and unrecognized model/config missing model_type are different validation failures." + }, + { + "left": "issue:44117", + "right": "issue:46328", + "accept": false, + "reason": "Tokenizer mapping fallback and SequenceFeatureExtractor.pad performance issue are unrelated." + }, + { + "left": "issue:43066", + "right": "issue:43122", + "accept": false, + "reason": "Both are tokenizer-related v5 regressions, but the evidence only supports a loose family resemblance; there is no clear shared root cause or single mergeable fix." + }, + { + "left": "issue:44964", + "right": "issue:45278", + "accept": false, + "reason": "A specific multimodal model-load failure and a broad import-error report are not the same bug." + }, + { + "left": "issue:44410", + "right": "issue:45478", + "accept": false, + "reason": "Different Qwen3.5 loading failures with different model details; not enough to treat as one duplicate." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall cluster summary is grounded: the packet does look heterogeneous, and using issue:45278 as a broad representative is reasonable. However, the soft-edge verdict for issue:43066 vs issue:43122 is too strong for the evidence in the packet. Both are tokenizer-related v5 regressions, but the titles only support a loose family resemblance, not a clearly shared root cause or a single mergeable bugfix. For this review, that positive accept is not conservative enough." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 47, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:38617", + "issue:41628", + "issue:42175", + "issue:42222", + "issue:42371", + "issue:42673", + "issue:42831", + "issue:43012", + "issue:43097", + "issue:43335", + "issue:43352", + "issue:43441", + "issue:43452", + "issue:43577", + "issue:43673", + "issue:43874", + "issue:43906", + "issue:43931", + "issue:43976", + "issue:43986", + "issue:44038", + "issue:44075", + "issue:44117", + "issue:44188", + "issue:44220", + "issue:44261", + "issue:44263", + "issue:44360", + "issue:44483", + "issue:44484", + "issue:44485", + "issue:44493", + "issue:44829", + "issue:44913", + "issue:44933", + "issue:44945", + "issue:44977", + "issue:44987", + "issue:45071", + "issue:45440", + "issue:45464", + "issue:45478", + "issue:45542", + "issue:45656", + "issue:46172", + "issue:46231", + "issue:46328" + ], + "soft_pairs": [ + "issue:42222|issue:45542", + "issue:42371|issue:44261", + "issue:44075|issue:45656", + "issue:44188|issue:46328", + "issue:38617|issue:44220", + "issue:44483|issue:45464", + "issue:43012|issue:43577", + "issue:42175|issue:42222", + "issue:44913|issue:46328", + "issue:43097|issue:44493", + "issue:44263|issue:44485", + "issue:43874|issue:44261", + "issue:41628|issue:45071", + "issue:45071|issue:45542", + "issue:27957|issue:43352", + "issue:44263|issue:46172", + "issue:44484|issue:45478", + "issue:44360|issue:46172", + "issue:43441|issue:44829", + "issue:42673|issue:44977", + "issue:42673|issue:43906", + "issue:42371|issue:42831", + "issue:43931|issue:44038", + "issue:43335|issue:45071", + "issue:43097|issue:43452", + "issue:44933|issue:46231", + "issue:43874|issue:44263", + "issue:43986|issue:44484", + "issue:43673|issue:46328", + "issue:43976|issue:44945", + "issue:44117|issue:44987", + "issue:45440|issue:46172" + ], + "prepared_review_unit_hash": "44bd56b631acc2f6e405709596e61c060d530ea61d5c92ce323f93cd336250b5", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 32, + "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 grab-bag of Transformers bug reports spanning unrelated areas (model-specific Flash Attention issues, import failures, backend/install problems, config regressions, and optimizer/runtime bugs). None of the soft-similarity pairs look like true duplicates of the same underlying bug or change.", + "confidence": 0.91, + "canonical_issue_reason": "issue:43097 is the best anchor for the set: it describes a concrete, user-visible v5 API break with downstream impact and is more representative of the version-regression theme than the narrower model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43097 is the strongest overall representative because it is a clear breaking-change report with broader applicability and higher triage value than the other, mostly isolated model/backend bugs.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:42222", + "right": "issue:45542", + "accept": false, + "reason": "Both involve TF/backend-related breakage, but one is a VitPose model failure and the other is a missing TensorFlow backend install path; different underlying bugs." + }, + { + "left": "issue:42371", + "right": "issue:44261", + "accept": false, + "reason": "TF32 settings guidance and an MLA RMS epsilon precision bug are unrelated precision topics, not the same defect." + }, + { + "left": "issue:44075", + "right": "issue:45656", + "accept": false, + "reason": "One is about SGD args being ignored, the other about optimizer.step() being called twice under DeepSpeed; different optimizer-path failures." + }, + { + "left": "issue:44188", + "right": "issue:46328", + "accept": false, + "reason": "Torch.compile attention-kernel divergence and SequenceFeatureExtractor.pad performance are unrelated code paths." + }, + { + "left": "issue:38617", + "right": "issue:44220", + "accept": false, + "reason": "An import error for configuration_utils and an audio feature extraction issue are different failures with no shared bug." + }, + { + "left": "issue:44483", + "right": "issue:45464", + "accept": false, + "reason": "Both affect chat/completions, but one is a v5.3 API rejection and the other is a Qwen3.5 streaming inference failure; not the same bug." + }, + { + "left": "issue:43012", + "right": "issue:43577", + "accept": false, + "reason": "A PyTorch warning when compiling bf16 models is not the same as Blip2 dtype propagation staying float32." + }, + { + "left": "issue:42175", + "right": "issue:42222", + "accept": false, + "reason": "Both mention TensorFlow, but one is packaging/backend installation and the other is model breakage in VitPose." + }, + { + "left": "issue:44913", + "right": "issue:46328", + "accept": false, + "reason": "GPTNeoX rotary_pct reload regression and a padding performance issue are unrelated." + }, + { + "left": "issue:43097", + "right": "issue:44493", + "accept": false, + "reason": "Removed tie_embeddings_and_encoder_decoder and unexpected position-id keys are different config/state-dict regressions." + }, + { + "left": "issue:44263", + "right": "issue:44485", + "accept": false, + "reason": "A torch.split return-value issue and a GLM-5 RoPE implementation issue are different tensor logic bugs." + }, + { + "left": "issue:43874", + "right": "issue:44261", + "accept": false, + "reason": "Missing get_number_of_image_patches and missing rms_norm_eps affect different multimodal/model components." + }, + { + "left": "issue:41628", + "right": "issue:45071", + "accept": false, + "reason": "AutoImageProcessor import failure and PreTrainedConfig type-checking regression are different API breakages." + }, + { + "left": "issue:45071", + "right": "issue:45542", + "accept": false, + "reason": "Type-checking regression in PretrainedConfig is unrelated to the TensorFlow backend/install issue." + }, + { + "left": "issue:27957", + "right": "issue:43352", + "accept": false, + "reason": "XLMRoberta Flash Attention 2 support and NemotronHForCausalLM Flash Attention 2 support are different model-specific capability gaps." + }, + { + "left": "issue:44263", + "right": "issue:46172", + "accept": false, + "reason": "Both touch tensor transforms, but the DSA indexer split bug and the deepseek-v4 transpose bug are different code paths." + }, + { + "left": "issue:44484", + "right": "issue:45478", + "accept": false, + "reason": "A shard-size question and a Qwen3.5-Moe from_pretrained error are unrelated." + }, + { + "left": "issue:44360", + "right": "issue:46172", + "accept": false, + "reason": "Missing ReLU in the DSA indexer and deepseek-v4 tensor transpose are distinct implementation bugs." + }, + { + "left": "issue:43441", + "right": "issue:44829", + "accept": false, + "reason": "Ministral FlashAttention failure and flash_attention_3 degenerate training are both attention-related but not the same defect." + }, + { + "left": "issue:42673", + "right": "issue:44977", + "accept": false, + "reason": "A VRAM leak in multithreaded loading is different from Qwen3.5 generation corruption under flash-attention." + }, + { + "left": "issue:42673", + "right": "issue:43906", + "accept": false, + "reason": "The VRAM leak and the isolated reproduction of issue 38071 are unrelated issues." + }, + { + "left": "issue:42371", + "right": "issue:42831", + "accept": false, + "reason": "TF32 behavior control and FineGrainedFP8 accuracy regression are separate precision/accuracy concerns." + }, + { + "left": "issue:43931", + "right": "issue:44038", + "accept": false, + "reason": "Qwen3-VL weight-shape mismatch is a loading mismatch, while Qwen3-VL-Moe on transformers 5.0 is a version-compatibility bug." + }, + { + "left": "issue:43335", + "right": "issue:45071", + "accept": false, + "reason": "A SwitchTransformers config edge case and a PretrainedConfig type-checking regression are different failures." + }, + { + "left": "issue:43097", + "right": "issue:43452", + "accept": false, + "reason": "Removal of tie_embeddings_and_encoder_decoder and gguf_file breaking tokenizer/model loading are unrelated loading/API issues." + }, + { + "left": "issue:44933", + "right": "issue:46231", + "accept": false, + "reason": "A nonexistent image_utils import and a documentation typo about set_training parameters are not the same bug." + }, + { + "left": "issue:43874", + "right": "issue:44263", + "accept": false, + "reason": "GLM46V image processor method absence and torch.split return values are unrelated component bugs." + }, + { + "left": "issue:43986", + "right": "issue:44484", + "accept": false, + "reason": "AutoProcessor crashing without torchvision and max_shard_size default behavior are unrelated." + }, + { + "left": "issue:43673", + "right": "issue:46328", + "accept": false, + "reason": "Generation cache missing during chunked prefill and SequenceFeatureExtractor.pad performance are different areas." + }, + { + "left": "issue:43976", + "right": "issue:44945", + "accept": false, + "reason": "Python version support for Transformers 5.1.0 and incorrect output under pipeline parallelism are unrelated regressions." + }, + { + "left": "issue:44117", + "right": "issue:44987", + "accept": false, + "reason": "TOKENIZER_MAPPING_NAMES returning None and loading physical-intelligence/fast failing are different loader issues." + }, + { + "left": "issue:45440", + "right": "issue:46172", + "accept": false, + "reason": "Native DeepseekV3MoE divergence from remote implementation and the deepseek-v4 comb.to/transpose bug are different DeepSeek-model issues, 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": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 48, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:38617", + "issue:42617", + "issue:42673", + "issue:42757", + "issue:42890", + "issue:42907", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43097", + "issue:43208", + "issue:43502", + "issue:43550", + "issue:43825", + "issue:43873", + "issue:43906", + "issue:43931", + "issue:43986", + "issue:43992", + "issue:44038", + "issue:44062", + "issue:44117", + "issue:44261", + "issue:44315", + "issue:44351", + "issue:44368", + "issue:44410", + "issue:44484", + "issue:44485", + "issue:44655", + "issue:44704", + "issue:44936", + "issue:44964", + "issue:45103", + "issue:45200", + "issue:45290", + "issue:45341", + "issue:45372", + "issue:45375", + "issue:45561", + "issue:45588", + "issue:45712", + "issue:45753", + "issue:45859", + "issue:46050", + "issue:46298", + "issue:46365", + "issue:46369" + ], + "soft_pairs": [ + "issue:43064|issue:44315", + "issue:38617|issue:42757", + "issue:43097|issue:45588", + "issue:43097|issue:46050", + "issue:43502|issue:43992", + "issue:44410|issue:45375", + "issue:42890|issue:46298", + "issue:44038|issue:44410", + "issue:43931|issue:44410", + "issue:42673|issue:45200", + "issue:38617|issue:44351", + "issue:44062|issue:44117", + "issue:44261|issue:44485", + "issue:45341|issue:45859", + "issue:43097|issue:43873", + "issue:44964|issue:45341", + "issue:44704|issue:45103", + "issue:43986|issue:44117", + "issue:45753|issue:46369", + "issue:42617|issue:45341", + "issue:43986|issue:45372", + "issue:43986|issue:45341", + "issue:45341|issue:45588", + "issue:45341|issue:45561", + "issue:43986|issue:45290", + "issue:43992|issue:44484", + "issue:44936|issue:46365", + "issue:43208|issue:43550", + "issue:43066|issue:43906", + "issue:43012|issue:43550", + "issue:42907|issue:44655", + "issue:43012|issue:44368", + "issue:45712|issue:46298", + "issue:43825|issue:44655", + "issue:42907|issue:45753", + "issue:43097|issue:46298" + ], + "prepared_review_unit_hash": "3fec7ee8ad0f8d948c4a70e3e62a5931a933c2da6515c794a33b0edeca33ea75", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38370, + "estimated_input_tokens": 9593, + "estimated_eval_tokens": 19442 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are largely unrelated regressions and edge-case bugs spread across different models, loading paths, and utility code. None of the soft pairs look like the same concrete bug/change, so I would not merge any of them as duplicates. Issue 43097 is the best cluster anchor because it is the broadest v5 regression report and has the strongest cross-cutting signal.", + "confidence": 0.86, + "canonical_issue_reason": "issue:43097 is the most umbrella-like report in the set: a Transformers v5 breaking-change regression with multiple inbound references and discussion, making it the best anchor for the cluster.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43097 is the most representative and broadly impactful issue here; it captures a cross-cutting v5 regression rather than a model-specific or one-off failure.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43064", + "right": "issue:44315", + "accept": false, + "reason": "Different failures: optimizer-state/rank mismatch in FSDP2+PEFT vs Liger Kernel not applied with model_init." + }, + { + "left": "issue:38617", + "right": "issue:42757", + "accept": false, + "reason": "Both are import errors, but for different missing symbols in different modules; not the same underlying bug." + }, + { + "left": "issue:43097", + "right": "issue:45588", + "accept": false, + "reason": "Unrelated code paths: removed v5 embedding API vs flash_attention crash on sink-less models." + }, + { + "left": "issue:43097", + "right": "issue:46050", + "accept": false, + "reason": "Different regression classes: removed v5 API surface vs quantized model save/load corruption." + }, + { + "left": "issue:43502", + "right": "issue:43992", + "accept": false, + "reason": "Both involve model loading, but one is unintended network access and the other is a missing state-dict weight." + }, + { + "left": "issue:44410", + "right": "issue:45375", + "accept": false, + "reason": "Both are Qwen config/model issues, but they concern different missing fields and different model variants." + }, + { + "left": "issue:42890", + "right": "issue:46298", + "accept": false, + "reason": "Test flakiness from missing seeding is unrelated to a noisy logging message." + }, + { + "left": "issue:44038", + "right": "issue:44410", + "accept": false, + "reason": "Different Qwen problems: a v5 compatibility bug for Qwen3-VL-Moe vs missing projections in qwen3next layer 0." + }, + { + "left": "issue:43931", + "right": "issue:44410", + "accept": false, + "reason": "Different model-loading failures: Qwen3-VL shape mismatch vs missing attention projections in a different model." + }, + { + "left": "issue:42673", + "right": "issue:45200", + "accept": false, + "reason": "VRAM leak in multithreaded dataloaders is unrelated to Gemma 4's mm_token_type_ids defaulting behavior." + }, + { + "left": "issue:38617", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different missing exports and different stack paths." + }, + { + "left": "issue:44062", + "right": "issue:44117", + "accept": false, + "reason": "AddedToken argument collision is a tokenizer construction bug; TOKENIZER_MAPPING_NAMES returning None is a separate mapping/lookup issue." + }, + { + "left": "issue:44261", + "right": "issue:44485", + "accept": false, + "reason": "Different model math issues: missing rms_norm_eps precision drift vs GLM-5 RoPE implementation." + }, + { + "left": "issue:45341", + "right": "issue:45859", + "accept": false, + "reason": "A testing_utils bug is unrelated to a CPU-offload incompatibility in Qwen3_5MoeTextRotaryEmbedding." + }, + { + "left": "issue:43097", + "right": "issue:43873", + "accept": false, + "reason": "One is a removed v5 embedding API; the other is an offloading/quantization behavior bug." + }, + { + "left": "issue:44964", + "right": "issue:45341", + "accept": false, + "reason": "Model loading for Phi-4 multimodal is unrelated to a testing_utils bug." + }, + { + "left": "issue:44704", + "right": "issue:45103", + "accept": false, + "reason": "Different layers entirely: AutoProcessor cached_file kwargs vs auto_docstring annotation parsing." + }, + { + "left": "issue:43986", + "right": "issue:44117", + "accept": false, + "reason": "Video AutoProcessor crash without torchvision is unrelated to tokenizer mapping returning None." + }, + { + "left": "issue:45753", + "right": "issue:46369", + "accept": false, + "reason": "Different symptoms and subsystems: infinite generation loop vs memory overflow in a different model." + }, + { + "left": "issue:42617", + "right": "issue:45341", + "accept": false, + "reason": "Cannot run 3d_parallel.py is unrelated to a testing_utils bug." + }, + { + "left": "issue:43986", + "right": "issue:45372", + "accept": false, + "reason": "Missing torchvision during video processor loading is unrelated to a mistral_common ReasoningEffort import failure." + }, + { + "left": "issue:43986", + "right": "issue:45290", + "accept": false, + "reason": "Processor loading without torchvision and chat-template tokenization on tool-call messages are different failures." + }, + { + "left": "issue:43992", + "right": "issue:44484", + "accept": false, + "reason": "Missing embed_tokens.weight during from_pretrained is unrelated to the default max_shard_size question." + }, + { + "left": "issue:44936", + "right": "issue:46365", + "accept": false, + "reason": "Trainer.evaluate after train is a runtime-behavior bug; Trainer import failing due to datasets is an environment/import issue." + }, + { + "left": "issue:43208", + "right": "issue:43550", + "accept": false, + "reason": "Different training/compile failures in different architectures; not the same code-path problem." + }, + { + "left": "issue:43066", + "right": "issue:43906", + "accept": false, + "reason": "Wrong tokenizer decoder type and an isolated reproduction of another issue are not the same concrete defect." + }, + { + "left": "issue:43012", + "right": "issue:43550", + "accept": false, + "reason": "A warning about bfloat16 compilation precision is not the same as a torch.compile SDPA failure." + }, + { + "left": "issue:42907", + "right": "issue:44655", + "accept": false, + "reason": "Saving dequantized models and saving Pipeline objects both involve save_pretrained, but they fail for different object types and paths." + }, + { + "left": "issue:43825", + "right": "issue:44655", + "accept": false, + "reason": "Incorrect translation-task messaging in pipeline() is unrelated to pipeline serialization failure." + }, + { + "left": "issue:45712", + "right": "issue:46298", + "accept": false, + "reason": "Left is dummy class cleanup/leakage, right is a logging-message complaint; no shared bug." + }, + { + "left": "issue:43097", + "right": "issue:46298", + "accept": false, + "reason": "Removed v5 embedding API and noisy logging are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 49, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36010", + "issue:36683", + "issue:38617", + "issue:42490", + "issue:42757", + "issue:42898", + "issue:42994", + "issue:43012", + "issue:43064", + "issue:43066", + "issue:43208", + "issue:43299", + "issue:43352", + "issue:43475", + "issue:43502", + "issue:43526", + "issue:43611", + "issue:43618", + "issue:43761", + "issue:43792", + "issue:43824", + "issue:43867", + "issue:43901", + "issue:43937", + "issue:44038", + "issue:44077", + "issue:44162", + "issue:44222", + "issue:44230", + "issue:44380", + "issue:44483", + "issue:44514", + "issue:44655", + "issue:44683", + "issue:44938", + "issue:45276", + "issue:45290", + "issue:45478", + "issue:45584", + "issue:45600", + "issue:45698", + "issue:46077", + "issue:46123", + "issue:46172", + "issue:46178", + "issue:46298", + "issue:46365", + "issue:46421" + ], + "soft_pairs": [ + "issue:44162|issue:46172", + "issue:44938|issue:46178", + "issue:43792|issue:46123", + "issue:44222|issue:45478", + "issue:42490|issue:43012", + "issue:44483|issue:44514", + "issue:43012|issue:44683", + "issue:43208|issue:43526", + "issue:38617|issue:43502", + "issue:42757|issue:43824", + "issue:45600|issue:46298", + "issue:44380|issue:46421", + "issue:45698|issue:46172", + "issue:44038|issue:44230", + "issue:44483|issue:45290", + "issue:43299|issue:43867", + "issue:43064|issue:44380", + "issue:36683|issue:43475", + "issue:42994|issue:44655", + "issue:43618|issue:43761", + "issue:42898|issue:43066", + "issue:45276|issue:46077", + "issue:43352|issue:43937", + "issue:36010|issue:42757", + "issue:43901|issue:44655", + "issue:43611|issue:44655", + "issue:45584|issue:46365", + "issue:44077|issue:46365" + ], + "prepared_review_unit_hash": "8b967c324909c1823cb900c6cdb8e23d46f782c84c7c31dd1ec235b21923588f", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 28, + "serialized_chars": 35643, + "estimated_input_tokens": 8911, + "estimated_eval_tokens": 18078 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This packet is still mostly a heterogeneous set of Transformers issues rather than one large duplicate cluster. The main representative issue is a broad open save/load regression (#45698), while the only soft edge that looks like the same underlying bug remains the pair of CLIP v5 output-regression issues.", + "confidence": 0.84, + "canonical_issue_reason": "No strong duplicate cluster emerges overall, so #45698 is a reasonable canonical issue because it is an open, user-facing save/load regression affecting custom module restoration. It is a useful representative anchor for triage, though it is not the most discussed issue in the packet.", + "canonical_pr_reason": null, + "best_issue_reason": "#45698 is the best single issue anchor: it is open, broadly impactful, and describes a substantive save/load regression. #36683 has higher discussion activity, but it is a narrower config-specific failure rather than the better representative issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44162", + "right": "issue:46172", + "accept": false, + "reason": "Different failures: ESM2 training breakage vs a DeepSeek-V4 tensor transpose/casting bug." + }, + { + "left": "issue:44938", + "right": "issue:46178", + "accept": false, + "reason": "Unrelated problems: Python 3.14 load failure vs a backend-mapping validation error for tensorflow_text." + }, + { + "left": "issue:43792", + "right": "issue:46123", + "accept": false, + "reason": "Whisper inference failure and MaskGenerationPipeline batch-dropping are separate code paths and symptoms." + }, + { + "left": "issue:44222", + "right": "issue:45478", + "accept": false, + "reason": "Both involve Qwen3.5/MoE FP8, but one is save_pretrained and the other is from_pretrained; not clearly the same concrete bug." + }, + { + "left": "issue:42490", + "right": "issue:43012", + "accept": false, + "reason": "A path-parsing import warning and a bf16 precision warning are unrelated regressions." + }, + { + "left": "issue:44483", + "right": "issue:44514", + "accept": false, + "reason": "Both touch chat APIs, but one is request acceptance at /v1/chat/completions and the other is a specific chat-template crash on batched input." + }, + { + "left": "issue:43012", + "right": "issue:44683", + "accept": false, + "reason": "Float32 precision warning during bf16 compilation is not the same as a compiled flex_attention failure on newer torch." + }, + { + "left": "issue:43208", + "right": "issue:43526", + "accept": false, + "reason": "xLSTM training bugs and BeitImageProcessorFast label reduction are unrelated model-component issues." + }, + { + "left": "issue:38617", + "right": "issue:43502", + "accept": false, + "reason": "An import error for layer_type_validation and offline-mode network requests are separate bugs." + }, + { + "left": "issue:42757", + "right": "issue:43824", + "accept": false, + "reason": "Both are import errors, but they concern different missing symbols and likely different package/export breakages." + }, + { + "left": "issue:45600", + "right": "issue:46298", + "accept": false, + "reason": "Broken auto_mappings references and a noisy logging message are not the same underlying issue." + }, + { + "left": "issue:44380", + "right": "issue:46421", + "accept": false, + "reason": "GPT2 attention-scaling behavior and cache lazy initialization recompilation are different backend behaviors." + }, + { + "left": "issue:45698", + "right": "issue:46172", + "accept": false, + "reason": "Custom module restoration after save_pretrained is unrelated to DeepSeek-V4 tensor conversion/transposition." + }, + { + "left": "issue:44038", + "right": "issue:44230", + "accept": false, + "reason": "Both mention Qwen3-VL/MoE, but one is a general Transformers 5 regression and the other is fp8 inference support." + }, + { + "left": "issue:44483", + "right": "issue:45290", + "accept": false, + "reason": "Both involve chat handling, but the reported failures are distinct and not clearly the same concrete bug." + }, + { + "left": "issue:43299", + "right": "issue:43867", + "accept": false, + "reason": "A base_model_prefix loading regression and a state_dict-sorting load error are different model-loading problems." + }, + { + "left": "issue:43064", + "right": "issue:44380", + "accept": false, + "reason": "Distributed training optimizer-state corruption and GPT2 attention scaling are unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43475", + "accept": false, + "reason": "Missing vocab_size on Gemma3Config and missing fpn_position_embeddings on Sam3VisionEncoderOutput are different attribute regressions." + }, + { + "left": "issue:42994", + "right": "issue:44655", + "accept": false, + "reason": "Quantized model save failure and pipeline save_pretrained failure are distinct save paths with different objects involved." + }, + { + "left": "issue:43618", + "right": "issue:43761", + "accept": true, + "reason": "Both look like the same CLIP v5 regression in output handling: one drops attentions and the other drops hidden_states when outputs are requested." + }, + { + "left": "issue:42898", + "right": "issue:43066", + "accept": false, + "reason": "Tokenizer cleanup behavior and wrong decoder type are separate v5 tokenizer issues." + }, + { + "left": "issue:45276", + "right": "issue:46077", + "accept": false, + "reason": "Gemma4 embedding resize behavior and bidirectional-attention masking are separate model-behavior bugs." + }, + { + "left": "issue:43352", + "right": "issue:43937", + "accept": false, + "reason": "Nemotron FlashAttention support and GLM-5 GenerationConfig validation are unrelated." + }, + { + "left": "issue:36010", + "right": "issue:42757", + "accept": false, + "reason": "Different missing imports from different modules; same general symptom but not the same bug." + }, + { + "left": "issue:43901", + "right": "issue:44655", + "accept": false, + "reason": "Documentation mismatch for TextClassificationPipeline is unrelated to pipeline serialization failure." + }, + { + "left": "issue:43611", + "right": "issue:44655", + "accept": false, + "reason": "Loading models with base_model_prefix and saving Pipeline objects are different regressions." + }, + { + "left": "issue:45584", + "right": "issue:46365", + "accept": false, + "reason": "Whisper empty-transcription generation and Trainer import interference from datasets are unrelated." + }, + { + "left": "issue:44077", + "right": "issue:46365", + "accept": false, + "reason": "An optional post_init contract issue in patchtsmixer is not the same as the Trainer import problem." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The overall clustering is mostly conservative, but the canonical issue rationale is not grounded: #45698 does not have the highest discussion activity in the packet (#36683 has 42 vs. #45698's 15). Please revise that claim and keep the representative-issue wording more cautious (e.g., a substantive open save/load regression), rather than asserting it as the most discussed issue." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 1, + "component_count": 20, + "review_unit_index": 50, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:28282", + "issue:33453", + "issue:36296", + "issue:36683", + "issue:41669", + "issue:42175", + "issue:42890", + "issue:43064", + "issue:43299", + "issue:43352", + "issue:43441", + "issue:43502", + "issue:43525", + "issue:43575", + "issue:43618", + "issue:43867", + "issue:43881", + "issue:43976", + "issue:43992", + "issue:44075", + "issue:44117", + "issue:44242", + "issue:44246", + "issue:44355", + "issue:44380", + "issue:44485", + "issue:44492", + "issue:44625", + "issue:44655", + "issue:44855", + "issue:44857", + "issue:44871", + "issue:44938", + "issue:44977", + "issue:45216", + "issue:45310", + "issue:45357", + "issue:45375", + "issue:45478", + "issue:45479", + "issue:45561", + "issue:45753", + "issue:46077", + "issue:46082", + "issue:46132", + "issue:46202", + "issue:46285", + "issue:46365" + ], + "soft_pairs": [ + "issue:44871|issue:46365", + "issue:43867|issue:45216", + "issue:43502|issue:44485", + "issue:44625|issue:46077", + "issue:46077|issue:46082", + "issue:36683|issue:43525", + "issue:43618|issue:45216", + "issue:42175|issue:46365", + "issue:44380|issue:46132", + "issue:36683|issue:44242", + "issue:36683|issue:43881", + "issue:43299|issue:45753", + "issue:43352|issue:43976", + "issue:43992|issue:45310", + "issue:44075|issue:46365", + "issue:43352|issue:45478", + "issue:44355|issue:44855", + "issue:43441|issue:44380", + "issue:43867|issue:45375", + "issue:43867|issue:46202", + "issue:36683|issue:44117", + "issue:43352|issue:44380", + "issue:44380|issue:45479", + "issue:41669|issue:44246", + "issue:42890|issue:45561", + "issue:43976|issue:44938", + "issue:43064|issue:44492", + "issue:43867|issue:45357", + "issue:43992|issue:45478", + "issue:36296|issue:43575", + "issue:43064|issue:44655", + "issue:28282|issue:33453", + "issue:43618|issue:46285", + "issue:44655|issue:44938", + "issue:43618|issue:44857", + "issue:43618|issue:44977" + ], + "prepared_review_unit_hash": "fb204f5856bd736eac935102c756e73b2141232227aa7383554bcbacbf9a292c", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 36, + "serialized_chars": 38140, + "estimated_input_tokens": 9535, + "estimated_eval_tokens": 19326 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of Transformers issues around loading/saving, config mismatches, attention backends, import/performance, and packaging. I don\u2019t see any soft pair that clearly reflects the same underlying bug/change, so I\u2019d keep them separate.", + "confidence": 0.82, + "canonical_issue_reason": "issue:43867 is the broadest user-facing model-loading failure in the set and the closest thing to a reusable umbrella anchor, though it is not a true duplicate of the others.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:43867 is the most generic and representative loading-regression report here, with a concrete reproduction and enough breadth to serve as the cluster\u2019s best issue anchor.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44871", + "right": "issue:46365", + "accept": false, + "reason": "Gemma eos-token inconsistency vs Trainer import conflict with datasets; unrelated bugs." + }, + { + "left": "issue:43867", + "right": "issue:45216", + "accept": false, + "reason": "Generic load error when state_dict is sorted vs Qwen3.5 save_pretrained checkpoint regression; different failure modes." + }, + { + "left": "issue:43502", + "right": "issue:44485", + "accept": false, + "reason": "local_files_only network leak vs GLM-5 RoPE discussion; not the same bug." + }, + { + "left": "issue:44625", + "right": "issue:46077", + "accept": false, + "reason": "Qwen3.5 num_labels propagation vs Gemma4 bidirectional attention-mask bug; different code paths." + }, + { + "left": "issue:46077", + "right": "issue:46082", + "accept": false, + "reason": "Gemma4 attention mask issue vs LlamaConfig head_dim validation; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43525", + "accept": false, + "reason": "Missing vocab_size vs missing pad_token_id; similar shape, different model-specific config defects." + }, + { + "left": "issue:43618", + "right": "issue:45216", + "accept": false, + "reason": "CLIP attentions API/structure change vs Qwen3.5 save_pretrained regression; different subsystems." + }, + { + "left": "issue:42175", + "right": "issue:46365", + "accept": false, + "reason": "Package/backend install problem vs Trainer import issue tied to datasets; not the same underlying bug." + }, + { + "left": "issue:44380", + "right": "issue:46132", + "accept": false, + "reason": "GPT2 scaling ignored in SDPA/FlashAttention vs AttentionInterface.register behavior change; distinct attention bugs." + }, + { + "left": "issue:36683", + "right": "issue:44242", + "accept": false, + "reason": "Gemma3 config attribute missing vs load-balancing loss omission; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:43881", + "accept": false, + "reason": "Gemma3 config attribute missing vs glm-4v-9b loading failure; different models and causes." + }, + { + "left": "issue:43299", + "right": "issue:45753", + "accept": false, + "reason": "Qwen3VL MoE loading break vs Qwen3_5 infinite-loop generation on one image; different symptoms." + }, + { + "left": "issue:43352", + "right": "issue:43976", + "accept": false, + "reason": "FlashAttention unsupported for Nemotron vs Python-version compatibility issue; unrelated." + }, + { + "left": "issue:43992", + "right": "issue:45310", + "accept": false, + "reason": "UMT5Encoder missing embed_tokens.weight vs Qwen3.5 MoE from_pretrained error; different models and failure mechanisms." + }, + { + "left": "issue:44075", + "right": "issue:46365", + "accept": false, + "reason": "SGD args ignored vs Trainer import problem; unrelated." + }, + { + "left": "issue:43352", + "right": "issue:45478", + "accept": false, + "reason": "Same broad FlashAttention area, but different model/repro and not clearly one fixable code-path bug." + }, + { + "left": "issue:44355", + "right": "issue:44855", + "accept": false, + "reason": "Compiled-file errors vs DebertaV2 import parse failure on Python 3.13; different root causes." + }, + { + "left": "issue:43441", + "right": "issue:44380", + "accept": false, + "reason": "Ministral FlashAttention failure vs GPT2 attention scaling bug; distinct model/backend issues." + }, + { + "left": "issue:43867", + "right": "issue:45375", + "accept": false, + "reason": "Generic load error vs Qwen3_5 vision config field being dropped; not the same defect." + }, + { + "left": "issue:43867", + "right": "issue:46202", + "accept": false, + "reason": "Model load failure vs Gemma pipeline roundtrip problem on \u2581; unrelated." + }, + { + "left": "issue:36683", + "right": "issue:44117", + "accept": false, + "reason": "Missing config attribute vs tokenizer-mapping returning None; different bug classes." + }, + { + "left": "issue:43352", + "right": "issue:44380", + "accept": false, + "reason": "FlashAttention support/error for one model vs GPT2 scaling being ignored; not the same issue." + }, + { + "left": "issue:44380", + "right": "issue:45479", + "accept": false, + "reason": "Attention scaling bug vs degenerate zero-loss classification; unrelated." + }, + { + "left": "issue:41669", + "right": "issue:44246", + "accept": false, + "reason": "Import-* cleanup/import-time regression vs intermittent import slowness; plausible theme overlap, but not the same concrete bug/change." + }, + { + "left": "issue:42890", + "right": "issue:45561", + "accept": false, + "reason": "Test seeding flakiness vs xdist race on captured_info.txt; separate test infra issues." + }, + { + "left": "issue:43976", + "right": "issue:44938", + "accept": false, + "reason": "Python version support mismatch vs Python 3.14 import failure; different compatibility problems." + }, + { + "left": "issue:43064", + "right": "issue:44492", + "accept": false, + "reason": "FSDP2/PEFT optimizer-state bug vs cache-strategy typo; unrelated." + }, + { + "left": "issue:43867", + "right": "issue:45357", + "accept": false, + "reason": "Model load error vs incorrect visual-encoder keys saved in Qwen3.5; different checkpoint problems." + }, + { + "left": "issue:43992", + "right": "issue:45478", + "accept": false, + "reason": "UMT5Encoder weight-missing issue vs Qwen3.5 MoE from_pretrained error; different model families and repros." + }, + { + "left": "issue:36296", + "right": "issue:43575", + "accept": false, + "reason": "Tensor parallel training bug vs tp loading OOM; one is a correctness bug, the other a memory-capacity issue." + }, + { + "left": "issue:43064", + "right": "issue:44655", + "accept": false, + "reason": "Optimizer-state mismatch during training vs pipeline save_pretrained failure; unrelated." + }, + { + "left": "issue:28282", + "right": "issue:33453", + "accept": false, + "reason": "Missing PyTorch dependency ImportError vs tokenizer-loading regression; not the same bug." + }, + { + "left": "issue:43618", + "right": "issue:46285", + "accept": false, + "reason": "CLIPOutput attentions removal vs CLIP text-model structure inconsistency; related family, but distinct changes." + }, + { + "left": "issue:44655", + "right": "issue:44938", + "accept": false, + "reason": "Pipeline save_pretrained bug vs Python 3.14 import failure; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44857", + "accept": false, + "reason": "CLIPOutput attentions API issue vs LwDetrImageLoss AMP/CUDA crash; unrelated." + }, + { + "left": "issue:43618", + "right": "issue:44977", + "accept": false, + "reason": "CLIP attentions regression vs Qwen3.5 flash-attention generation loop; 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": 20, + "review_unit_index": 51, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:27957", + "issue:30333", + "issue:30990", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:42548", + "issue:42757", + "issue:42831", + "issue:42890", + "issue:43010", + "issue:43299", + "issue:43502", + "issue:43576", + "issue:43723", + "issue:43824", + "issue:43827", + "issue:43874", + "issue:43976", + "issue:44062", + "issue:44112", + "issue:44222", + "issue:44315", + "issue:44351", + "issue:44355", + "issue:44360", + "issue:44368", + "issue:44387", + "issue:44556", + "issue:44655", + "issue:44749", + "issue:44908", + "issue:44938", + "issue:44961", + "issue:44964", + "issue:45071", + "issue:45125", + "issue:45278", + "issue:45375", + "issue:45440", + "issue:45468", + "issue:45478", + "issue:45600", + "issue:45656", + "issue:45850", + "issue:46132" + ], + "soft_pairs": [ + "issue:44360|issue:45440", + "issue:43502|issue:44062", + "issue:42890|issue:44964", + "issue:42890|issue:44112", + "issue:42890|issue:45600", + "issue:44222|issue:45125", + "issue:43576|issue:46132", + "issue:45125|issue:45478", + "issue:42548|issue:44351", + "issue:44355|issue:44749", + "issue:30333|issue:33453", + "issue:44387|issue:44556", + "issue:45125|issue:45375", + "issue:43010|issue:45278", + "issue:43723|issue:44556", + "issue:30990|issue:31795", + "issue:43502|issue:44222", + "issue:43502|issue:44961", + "issue:42831|issue:43874", + "issue:43010|issue:45850", + "issue:30333|issue:36296", + "issue:44355|issue:45071", + "issue:43010|issue:44368", + "issue:43827|issue:44961", + "issue:44355|issue:44655", + "issue:33453|issue:34689", + "issue:44315|issue:44360", + "issue:43299|issue:44556", + "issue:27957|issue:33453", + "issue:42757|issue:43576", + "issue:43576|issue:45071", + "issue:43010|issue:45656", + "issue:43576|issue:43824", + "issue:31795|issue:38617", + "issue:44961|issue:45468", + "issue:43010|issue:43976", + "issue:43576|issue:44938", + "issue:44908|issue:44961" + ], + "prepared_review_unit_hash": "3e260e6649eb61047cda00e84edb951ee4761afa2723d04a80ef6818cde5a7bc", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 38, + "serialized_chars": 38630, + "estimated_input_tokens": 9658, + "estimated_eval_tokens": 19572 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items are mostly unrelated issue reports; the only recurring pattern is broad v5 regressions/import/load failures, but none of the soft pairs look like the same underlying bug.", + "confidence": 0.84, + "canonical_issue_reason": "issue:45278 is the clearest umbrella-style report here: it clearly states a broad version-regression symptom ('many import errors after update from 4.57.0 to 5.5.0') and is more representative than the narrowly scoped one-off bugs.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the best single issue to anchor the cluster because it is specific enough to be understandable yet broad enough to represent the shared v5 breakage theme seen in nearby reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44360", + "right": "issue:45440", + "accept": false, + "reason": "Completely different problems: a DSA indexer/ReLU discussion vs a DeepSeekV3MoE implementation divergence." + }, + { + "left": "issue:43502", + "right": "issue:44062", + "accept": false, + "reason": "One is about unwanted network access with local_files_only=True; the other is a tokenizers.AddedToken argument error." + }, + { + "left": "issue:42890", + "right": "issue:44964", + "accept": false, + "reason": "Seed-related flaky tests for SamHQ are unrelated to Phi-4 multimodal model loading failures." + }, + { + "left": "issue:42890", + "right": "issue:44112", + "accept": false, + "reason": "A missing set_seed() in a test is unrelated to a GraniteSpeech CI device-override failure." + }, + { + "left": "issue:42890", + "right": "issue:45600", + "accept": false, + "reason": "Test flakiness in SamHQ does not match an auto_mappings.py CI break from removed configs." + }, + { + "left": "issue:44222", + "right": "issue:45125", + "accept": false, + "reason": "Both mention MoE/FP8-adjacent areas, but one is about save_pretrained with FP8 and the other about missing _tp_plan for tensor parallelism." + }, + { + "left": "issue:43576", + "right": "issue:46132", + "accept": false, + "reason": "A broken CLI env command and AttentionInterface.register behavior changes are different code paths and symptoms." + }, + { + "left": "issue:45125", + "right": "issue:45478", + "accept": false, + "reason": "Same model family, but one is missing tensor-parallel metadata while the other is a from_pretrained regression; not the same concrete bug." + }, + { + "left": "issue:42548", + "right": "issue:44351", + "accept": false, + "reason": "Both are import errors, but they involve different symbols and likely different breakages, so they are not the same bug." + }, + { + "left": "issue:44355", + "right": "issue:44749", + "accept": false, + "reason": "Running compiled Python files errors and slowed filtering after upgrade are unrelated." + }, + { + "left": "issue:30333", + "right": "issue:33453", + "accept": false, + "reason": "MLflow reporting on failed training jobs is unrelated to tokenizer-loading regression." + }, + { + "left": "issue:44387", + "right": "issue:44556", + "accept": false, + "reason": "Memory growth under int4 quantization is a different failure mode from checkpoint reload incompatibility." + }, + { + "left": "issue:45125", + "right": "issue:45375", + "accept": false, + "reason": "Both are Qwen3.5 MoE-related, but one is missing tensor-parallel plan data while the other is a missing vision config field silently dropped by strict parsing." + }, + { + "left": "issue:43010", + "right": "issue:45278", + "accept": false, + "reason": "A no_grad decoration request for cache updates is unrelated to import errors after upgrading to v5.5.0." + }, + { + "left": "issue:43723", + "right": "issue:44556", + "accept": false, + "reason": "Tokenizer loading issues and checkpoint reload failures are both load-related but not the same underlying problem." + }, + { + "left": "issue:30990", + "right": "issue:31795", + "accept": false, + "reason": "Sentence Transformers loading hangs and documentation confusion about model.forward arguments are unrelated." + }, + { + "left": "issue:43502", + "right": "issue:44222", + "accept": false, + "reason": "Local-files-only network leakage is unrelated to FP8 save_pretrained behavior." + }, + { + "left": "issue:43502", + "right": "issue:44961", + "accept": false, + "reason": "No meaningful overlap: offline-loading network calls vs an unrelated placeholder issue." + }, + { + "left": "issue:42831", + "right": "issue:43874", + "accept": false, + "reason": "An accuracy issue in FineGrainedFP8 is not the same as a missing image-patch helper causing an AttributeError." + }, + { + "left": "issue:43010", + "right": "issue:45850", + "accept": false, + "reason": "A cache update no_grad change is unrelated to a vague general complaint that v5 breaks too many models." + }, + { + "left": "issue:30333", + "right": "issue:36296", + "accept": false, + "reason": "MLflow reporting and tensor-parallel training bugs are different subsystems and failure modes." + }, + { + "left": "issue:44355", + "right": "issue:45071", + "accept": false, + "reason": "Compiled-file execution errors are unrelated to PretrainedConfig type-checking regressions." + }, + { + "left": "issue:43010", + "right": "issue:44368", + "accept": false, + "reason": "Cache update semantics have nothing to do with a tie_word_embeddings warning in Qwen3.5 LoRA fine-tuning." + }, + { + "left": "issue:43827", + "right": "issue:44961", + "accept": false, + "reason": "Docs still referencing pipeline() and an unrelated placeholder issue are not duplicates." + }, + { + "left": "issue:44355", + "right": "issue:44655", + "accept": false, + "reason": "Compiled Python file errors and inability to save Pipeline objects are different issues." + }, + { + "left": "issue:33453", + "right": "issue:34689", + "accept": false, + "reason": "Tokenizer-loading regression and Llama 3.2 model-loading breakage are both loading failures, but they are not the same concrete bug." + }, + { + "left": "issue:44315", + "right": "issue:44360", + "accept": false, + "reason": "Liger Kernel application with model_init is unrelated to a DSA indexer missing ReLU." + }, + { + "left": "issue:43299", + "right": "issue:44556", + "accept": false, + "reason": "Qwen3VL MoE loading under dev0 and checkpoint reload after upgrading to v5.2/v5.3 are distinct regression reports." + }, + { + "left": "issue:27957", + "right": "issue:33453", + "accept": false, + "reason": "Flash Attention 2 support for XLMRoberta is unrelated to tokenizer loading regression." + }, + { + "left": "issue:42757", + "right": "issue:43576", + "accept": false, + "reason": "An import error from huggingface_hub is not the same as the transformers env CLI being broken." + }, + { + "left": "issue:43576", + "right": "issue:45071", + "accept": false, + "reason": "Broken env CLI behavior and PretrainedConfig type-checking are unrelated." + }, + { + "left": "issue:43010", + "right": "issue:45656", + "accept": false, + "reason": "Cache no_grad decoration and double optimizer steps under deepspeed are different training-path bugs." + }, + { + "left": "issue:43576", + "right": "issue:43824", + "accept": false, + "reason": "The env command failure is unrelated to Qwen2.5-VL import failures." + }, + { + "left": "issue:31795", + "right": "issue:38617", + "accept": false, + "reason": "Documentation confusion and an ImportError for layer_type_validation are unrelated." + }, + { + "left": "issue:44961", + "right": "issue:45468", + "accept": false, + "reason": "An unrelated placeholder issue does not match the Gemma-4 audio positional encoding bug." + }, + { + "left": "issue:43010", + "right": "issue:43976", + "accept": false, + "reason": "Cache update no_grad behavior is unrelated to Python 3.9/3.10 support problems." + }, + { + "left": "issue:43576", + "right": "issue:44938", + "accept": false, + "reason": "The transformers env command and Python 3.14 load failure are different issues." + }, + { + "left": "issue:44908", + "right": "issue:44961", + "accept": false, + "reason": "Scheduler kwargs handling has no relation to the unrelated 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": 20, + "review_unit_index": 52, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:36296", + "issue:41628", + "issue:41669", + "issue:42548", + "issue:42890", + "issue:43352", + "issue:43408", + "issue:43525", + "issue:43526", + "issue:43572", + "issue:43576", + "issue:43673", + "issue:43874", + "issue:43901", + "issue:43976", + "issue:44038", + "issue:44222", + "issue:44230", + "issue:44246", + "issue:44273", + "issue:44295", + "issue:44297", + "issue:44315", + "issue:44351", + "issue:44373", + "issue:44393", + "issue:44556", + "issue:44623", + "issue:44704", + "issue:44908", + "issue:44961", + "issue:45003", + "issue:45071", + "issue:45092", + "issue:45125", + "issue:45230", + "issue:45278", + "issue:45341", + "issue:45362", + "issue:45478", + "issue:45542", + "issue:45753", + "issue:45850", + "issue:46095", + "issue:46143", + "issue:46144", + "issue:46209", + "issue:46218" + ], + "soft_pairs": [ + "issue:44222|issue:44230", + "issue:36296|issue:43901", + "issue:41628|issue:42548", + "issue:44315|issue:46209", + "issue:36296|issue:43526", + "issue:44623|issue:45230", + "issue:44273|issue:45542", + "issue:44038|issue:45125", + "issue:44393|issue:45071", + "issue:44273|issue:45003", + "issue:44373|issue:44961", + "issue:44230|issue:46095", + "issue:44393|issue:45753", + "issue:43576|issue:44556", + "issue:43525|issue:45850", + "issue:41669|issue:43673", + "issue:44556|issue:45092", + "issue:43572|issue:45850", + "issue:45362|issue:45850", + "issue:41669|issue:46218", + "issue:43408|issue:44315", + "issue:42890|issue:44908", + "issue:43352|issue:44393", + "issue:44246|issue:44273", + "issue:44297|issue:45230", + "issue:44295|issue:45230", + "issue:43576|issue:44908", + "issue:43976|issue:44393", + "issue:44908|issue:45071", + "issue:44704|issue:44908", + "issue:41669|issue:44351", + "issue:45278|issue:45850", + "issue:41669|issue:45341", + "issue:44393|issue:45478", + "issue:43874|issue:45850", + "issue:46143|issue:46144", + "issue:43976|issue:46218" + ], + "prepared_review_unit_hash": "83a279587a3c88f19e45bf5f3a45a09d3cd416b5bc11d4d9649e7b6cb52b17d9", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 48, + "item_count": 48, + "soft_pair_count": 37, + "serialized_chars": 38060, + "estimated_input_tokens": 9515, + "estimated_eval_tokens": 19286 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "Mostly a noisy mix of unrelated Transformers v5 regression reports, with a few broad import/lazy-loading complaints and many model-specific bugs. The soft pairs share only superficial subsystem overlap, not the same underlying defect or change.", + "confidence": 0.86, + "canonical_issue_reason": "issue:45278 is the best umbrella issue here: it explicitly captures the broad \"many import errors after update from 4.57.0 to 5.5.0\" regression theme that most closely matches the cluster\u2019s dominant pattern.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45278 is the most representative of the cluster because it generalizes the repeated import/load breakages without being as vague as the catch-all complaints.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44222", + "right": "issue:44230", + "accept": false, + "reason": "Both involve fp8/moe, but one is about save_pretrained behavior and the other about inference support; different code paths and fixes." + }, + { + "left": "issue:36296", + "right": "issue:43901", + "accept": false, + "reason": "Tensor parallel training bug vs a pipeline doc mismatch; no shared underlying bug." + }, + { + "left": "issue:41628", + "right": "issue:42548", + "accept": false, + "reason": "Both are import errors, but for different exported symbols and likely different breakages." + }, + { + "left": "issue:44315", + "right": "issue:46209", + "accept": false, + "reason": "Liger Kernel application in model_init is unrelated to FP8 inference failure on B200." + }, + { + "left": "issue:36296", + "right": "issue:43526", + "accept": false, + "reason": "Tensor parallel training vs BeitImageProcessorFast label reduction; unrelated subsystems and symptoms." + }, + { + "left": "issue:44623", + "right": "issue:45230", + "accept": false, + "reason": "A processor save_pretrained file omission is not the same as a vague general bug report." + }, + { + "left": "issue:44273", + "right": "issue:45542", + "accept": false, + "reason": "Lazy loading problems and tensorboard/TF backend detection are different bugs." + }, + { + "left": "issue:44038", + "right": "issue:45125", + "accept": false, + "reason": "Qwen3-VL-Moe bug vs missing _tp_plan for tensor parallelism; related model family but different concrete failures." + }, + { + "left": "issue:44393", + "right": "issue:45071", + "accept": false, + "reason": "Qwen3-VL 2D box output issue vs PretrainedConfig type checking regression; not the same defect." + }, + { + "left": "issue:44273", + "right": "issue:45003", + "accept": false, + "reason": "Lazy loading behavior and sys.modules access safety are distinct implementation problems." + }, + { + "left": "issue:44373", + "right": "issue:44961", + "accept": false, + "reason": "Docstring wording for position_ids vs an unrelated one-word issue title; no evidence of shared bug." + }, + { + "left": "issue:44230", + "right": "issue:46095", + "accept": false, + "reason": "FP8 inference support for Qwen models is not the same as asking about a weight conversion script." + }, + { + "left": "issue:44393", + "right": "issue:45753", + "accept": false, + "reason": "Both involve Qwen3.5-VL behavior, but one is 2D box output/hallucination and the other is an infinite loop on a specific image; too different to merge." + }, + { + "left": "issue:43576", + "right": "issue:44556", + "accept": false, + "reason": "Broken `transformers env` command and checkpoint reload incompatibility after upgrade are separate issues." + }, + { + "left": "issue:43525", + "right": "issue:45850", + "accept": false, + "reason": "A missing pad_token_id on Llama4Config is a specific config bug, while the other issue is a broad complaint about many models breaking." + }, + { + "left": "issue:41669", + "right": "issue:43673", + "accept": false, + "reason": "Import * usage slowing imports is not the same as missing GenerationMixin cache during chunked prefill." + }, + { + "left": "issue:44556", + "right": "issue:45092", + "accept": false, + "reason": "Both are upgrade/meta-init related, but one is checkpoint reload incompatibility and the other is remote-code checkpoint incompatibility; different concrete failures." + }, + { + "left": "issue:43572", + "right": "issue:45850", + "accept": false, + "reason": "Missing pad_token_idx in one config is a specific regression; the other is a broad multi-model complaint." + }, + { + "left": "issue:45362", + "right": "issue:45850", + "accept": false, + "reason": "Qwen3.5-35B chat crash is model-specific, not the same as a generalized v5 breakage report." + }, + { + "left": "issue:41669", + "right": "issue:46218", + "accept": false, + "reason": "Import * usage from models and inability to import Qwen3_5ForCausalLM are both import-related, but not the same bug or fix." + }, + { + "left": "issue:43408", + "right": "issue:44315", + "accept": false, + "reason": "sam3_video/sam3_tracker warning is a model-type mismatch, unrelated to Liger Kernel application in model_init." + }, + { + "left": "issue:42890", + "right": "issue:44908", + "accept": false, + "reason": "Missing set_seed in a SAM HQ integration test is unrelated to inverse_sqrt scheduler kwargs handling." + }, + { + "left": "issue:43352", + "right": "issue:44393", + "accept": false, + "reason": "Flash Attention 2 support error for NemotronH vs Qwen3-VL 2D box output failure are different model and failure modes." + }, + { + "left": "issue:44246", + "right": "issue:44273", + "accept": false, + "reason": "Import sometimes taking long and lazy loading not working are related only at a very broad loading-performance level, not the same underlying defect." + }, + { + "left": "issue:44297", + "right": "issue:45230", + "accept": false, + "reason": "Tokenizer class metadata mismatch on save_pretrained is distinct from a generic bug report." + }, + { + "left": "issue:44295", + "right": "issue:45230", + "accept": false, + "reason": "Reading position_ids after buffer registration is a specific tensor/buffer issue, not the same as a generic bug report." + }, + { + "left": "issue:43576", + "right": "issue:44908", + "accept": false, + "reason": "Environment command breakage and scheduler kwargs ignored are unrelated features." + }, + { + "left": "issue:43976", + "right": "issue:44393", + "accept": false, + "reason": "Python version compatibility for transformers 5.1.0 is unrelated to a Qwen3-VL output error." + }, + { + "left": "issue:44908", + "right": "issue:45071", + "accept": false, + "reason": "Scheduler kwargs handling and PretrainedConfig type checking are separate issues." + }, + { + "left": "issue:44704", + "right": "issue:44908", + "accept": false, + "reason": "AutoProcessor.cached_file kwargs propagation and scheduler kwargs handling are different code paths." + }, + { + "left": "issue:41669", + "right": "issue:44351", + "accept": false, + "reason": "import * usage removal from models and missing HybridCache import are both import-related, but not the same defect." + }, + { + "left": "issue:45278", + "right": "issue:45850", + "accept": false, + "reason": "A concrete set of import errors after upgrade is not the same as a vague broad complaint that v5 breaks too many models." + }, + { + "left": "issue:41669", + "right": "issue:45341", + "accept": false, + "reason": "import * performance/import issue and a testing_utils bug are unrelated." + }, + { + "left": "issue:44393", + "right": "issue:45478", + "accept": false, + "reason": "Qwen3-VL 2D box hallucination/error and Qwen3.5 Moe from_pretrained error are different model failures." + }, + { + "left": "issue:43874", + "right": "issue:45850", + "accept": false, + "reason": "Missing get_number_of_image_patches in one image processor is a specific API omission, not the same as a broad v5 complaint." + }, + { + "left": "issue:46143", + "right": "issue:46144", + "accept": false, + "reason": "Not passing kwargs through RoFormer methods and attention implementation not using the attention interface are related to RoFormer, but they are different code-path problems." + }, + { + "left": "issue:43976", + "right": "issue:46218", + "accept": false, + "reason": "Python version support regression is unrelated to a module import failure for Qwen3_5ForCausalLM." + } + ] + }, + "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": 20, + "review_unit_index": 53, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:30333", + "issue:31515", + "issue:31795", + "issue:33453", + "issue:34689", + "issue:36296", + "issue:38617", + "issue:41669", + "issue:42371", + "issue:42913", + "issue:43066", + "issue:43298", + "issue:43352", + "issue:43502", + "issue:43519", + "issue:43525", + "issue:43572", + "issue:43576", + "issue:43606", + "issue:43673", + "issue:43704", + "issue:43827", + "issue:44077", + "issue:44230", + "issue:44261", + "issue:44263", + "issue:44273", + "issue:44315", + "issue:44351", + "issue:44360", + "issue:44485", + "issue:44492", + "issue:44623", + "issue:44655", + "issue:44829", + "issue:44861", + "issue:44908", + "issue:44998", + "issue:45230", + "issue:45290", + "issue:45310", + "issue:45468", + "issue:45850", + "issue:46018", + "issue:46095", + "issue:46172", + "issue:46218" + ], + "soft_pairs": [ + "issue:44077|issue:44861", + "issue:43673|issue:44492", + "issue:43827|issue:44908", + "issue:31795|issue:33453", + "issue:41669|issue:44492", + "issue:44315|issue:44829", + "issue:43576|issue:44273", + "issue:44655|issue:44861", + "issue:44360|issue:45468", + "issue:45230|issue:45310", + "issue:42371|issue:44492", + "issue:44273|issue:44908", + "issue:43576|issue:46218", + "issue:43827|issue:44492", + "issue:43519|issue:44485", + "issue:43519|issue:44908", + "issue:46018|issue:46172", + "issue:41669|issue:43704", + "issue:31795|issue:34689", + "issue:31515|issue:31795", + "issue:43704|issue:44485", + "issue:44485|issue:45468", + "issue:45850|issue:46218", + "issue:42371|issue:43704", + "issue:43298|issue:46095", + "issue:43502|issue:43519", + "issue:43519|issue:44623", + "issue:43519|issue:43704", + "issue:44263|issue:45468", + "issue:43606|issue:45468", + "issue:44261|issue:45468", + "issue:44351|issue:46218", + "issue:43352|issue:46218", + "issue:42913|issue:43066", + "issue:44908|issue:44998", + "issue:44861|issue:45290", + "issue:43704|issue:44998", + "issue:43525|issue:44861", + "issue:31515|issue:38617", + "issue:43572|issue:44861", + "issue:43066|issue:44998", + "issue:41669|issue:44998", + "issue:44998|issue:46218", + "issue:44230|issue:44623", + "issue:30333|issue:31515", + "issue:31515|issue:36296", + "issue:44230|issue:45310", + "issue:31515|issue:33453", + "issue:43298|issue:44230" + ], + "prepared_review_unit_hash": "0a5267d7f3f7e4397c6f64edfa008ea6e990570eded90500ebee3e5bc0521f79", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 47, + "item_count": 47, + "soft_pair_count": 49, + "serialized_chars": 41024, + "estimated_input_tokens": 10256, + "estimated_eval_tokens": 20768 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "These items do not form a real duplicate cluster; the soft-similarity links mostly connect unrelated issues that only share broad vocabulary (v5, loading, cache, tokenizer, etc.). I\u2019d keep one concrete, broadly scoped issue as the representative anchor and reject all soft edges.", + "confidence": 0.74, + "canonical_issue_reason": "No true canonical duplicate set emerges here; issue 41669 is the strongest standalone anchor because it has a concrete, broadly impactful regression with clearer scope and more discussion than the vague or highly model-specific reports.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 41669 is the best single issue to retain as a representative reference point: it is specific, actionable, and has broader ecosystem impact than the other mostly narrow or ambiguous reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44077", + "right": "issue:44861", + "accept": false, + "reason": "Different bugs: post_init validation vs AttributeError in tied-weight key handling." + }, + { + "left": "issue:43673", + "right": "issue:44492", + "accept": false, + "reason": "Chunked prefill cache behavior is unrelated to a typo in cache strategy docs." + }, + { + "left": "issue:43827", + "right": "issue:44908", + "accept": false, + "reason": "Documentation cleanup after pipeline removals is unrelated to scheduler kwargs handling." + }, + { + "left": "issue:31795", + "right": "issue:33453", + "accept": false, + "reason": "Doc confusion about forward args is not the same as a tokenizer loading regression." + }, + { + "left": "issue:41669", + "right": "issue:44492", + "accept": false, + "reason": "Import-star performance issue is unrelated to a cache-strategy typo." + }, + { + "left": "issue:44315", + "right": "issue:44829", + "accept": false, + "reason": "Model-init Liger Kernel application and flash-attention training degeneration are different failures." + }, + { + "left": "issue:43576", + "right": "issue:44273", + "accept": false, + "reason": "Broken env command is unrelated to lazy loading behavior." + }, + { + "left": "issue:44655", + "right": "issue:44861", + "accept": false, + "reason": "Pipeline save_pretrained support is unrelated to tied-weight key crashes." + }, + { + "left": "issue:44360", + "right": "issue:45468", + "accept": false, + "reason": "DSA indexer ReLU discussion is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:45230", + "right": "issue:45310", + "accept": false, + "reason": "Generic bug report placeholder vs a specific Qwen3.5 MoE from_pretrained error." + }, + { + "left": "issue:42371", + "right": "issue:44492", + "accept": false, + "reason": "TF32 API migration is unrelated to a cache-strategy typo." + }, + { + "left": "issue:44273", + "right": "issue:44908", + "accept": false, + "reason": "Lazy loading and inverse_sqrt scheduler kwargs are different code paths." + }, + { + "left": "issue:43576", + "right": "issue:46218", + "accept": false, + "reason": "Env command breakage is unrelated to a Qwen3_5 import failure." + }, + { + "left": "issue:43827", + "right": "issue:44492", + "accept": false, + "reason": "Pipeline-removal docs issue is unrelated to a cache-strategy typo." + }, + { + "left": "issue:43519", + "right": "issue:44485", + "accept": false, + "reason": "Timestamp calculation in Qwen3VL is unrelated to GLM-5 RoPE implementation." + }, + { + "left": "issue:43519", + "right": "issue:44908", + "accept": false, + "reason": "Timestamp logic and scheduler kwargs are unrelated bugs." + }, + { + "left": "issue:46018", + "right": "issue:46172", + "accept": false, + "reason": "DeepSeek shared-expert gating and a tensor transpose issue are different model bugs." + }, + { + "left": "issue:41669", + "right": "issue:43704", + "accept": false, + "reason": "Import-star removal performance work is unrelated to a VRAM leak in dataloader threads." + }, + { + "left": "issue:31795", + "right": "issue:34689", + "accept": false, + "reason": "Forward-doc confusion is unrelated to Llama 3.2 model loading regression." + }, + { + "left": "issue:31515", + "right": "issue:31795", + "accept": false, + "reason": "Slow checkpoint loading is not the same as forward-doc confusion." + }, + { + "left": "issue:43704", + "right": "issue:44485", + "accept": false, + "reason": "VRAM leak in threaded loading is unrelated to GLM-5 RoPE behavior." + }, + { + "left": "issue:44485", + "right": "issue:45468", + "accept": false, + "reason": "Both are model-specific math issues, but they affect different architectures and code paths." + }, + { + "left": "issue:45850", + "right": "issue:46218", + "accept": false, + "reason": "Broad v5 breakage report is not the same as a specific Qwen3_5 import error." + }, + { + "left": "issue:42371", + "right": "issue:43704", + "accept": false, + "reason": "TF32 settings API and VRAM leak reports are unrelated." + }, + { + "left": "issue:43298", + "right": "issue:46095", + "accept": false, + "reason": "Continuous batching support for audio models is unrelated to DeepSeek-V4 weight conversion." + }, + { + "left": "issue:43502", + "right": "issue:43519", + "accept": false, + "reason": "local_files_only network leakage is unrelated to Qwen3VL timestamp computation." + }, + { + "left": "issue:43519", + "right": "issue:44623", + "accept": false, + "reason": "Timestamp math and missing processor.save_pretrained files are different bugs." + }, + { + "left": "issue:43519", + "right": "issue:43704", + "accept": false, + "reason": "Timestamp calculation bug is unrelated to VRAM leakage in threaded loading." + }, + { + "left": "issue:44263", + "right": "issue:45468", + "accept": false, + "reason": "torch.split return-value concerns are unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:43606", + "right": "issue:45468", + "accept": false, + "reason": "CPU offload device mismatch is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44261", + "right": "issue:45468", + "accept": false, + "reason": "RMS norm epsilon precision mismatch is unrelated to Gemma audio positional encoding." + }, + { + "left": "issue:44351", + "right": "issue:46218", + "accept": false, + "reason": "HybridCache import failure and Qwen3_5 import failure are different symbols and likely different fixes." + }, + { + "left": "issue:43352", + "right": "issue:46218", + "accept": false, + "reason": "Flash Attention 2.0 support gating is unrelated to an import error." + }, + { + "left": "issue:42913", + "right": "issue:43066", + "accept": false, + "reason": "Both mention tokenizer v5, but one is broad behavior drift and the other is a specific decoder-type issue." + }, + { + "left": "issue:44908", + "right": "issue:44998", + "accept": false, + "reason": "Scheduler kwargs bug is unrelated to the placeholder 'Unemployment' issue." + }, + { + "left": "issue:44861", + "right": "issue:45290", + "accept": false, + "reason": "Tied-weight key handling and chat-template tool-call crash are different failures." + }, + { + "left": "issue:43704", + "right": "issue:44998", + "accept": false, + "reason": "VRAM leak issue is unrelated to the placeholder report." + }, + { + "left": "issue:43525", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_id in Llama4Config is unrelated to tied-weight key crashes." + }, + { + "left": "issue:31515", + "right": "issue:38617", + "accept": false, + "reason": "Slow from_pretrained checkpoint loading is unrelated to an ImportError for layer_type_validation." + }, + { + "left": "issue:43572", + "right": "issue:44861", + "accept": false, + "reason": "Missing pad_token_idx after v5 update is unrelated to tied-weight key crashes." + }, + { + "left": "issue:43066", + "right": "issue:44998", + "accept": false, + "reason": "Tokenizer decoder type bug is unrelated to the placeholder report." + }, + { + "left": "issue:41669", + "right": "issue:44998", + "accept": false, + "reason": "Import-star removal/performance issue is unrelated to the placeholder report." + }, + { + "left": "issue:44998", + "right": "issue:46218", + "accept": false, + "reason": "Placeholder issue is unrelated to a specific Qwen3_5 import failure." + }, + { + "left": "issue:44230", + "right": "issue:44623", + "accept": false, + "reason": "FP8 support work is unrelated to missing files in processor.save_pretrained." + }, + { + "left": "issue:30333", + "right": "issue:31515", + "accept": false, + "reason": "MLFlow reporting on failed training jobs is unrelated to slow checkpoint loading." + }, + { + "left": "issue:31515", + "right": "issue:36296", + "accept": false, + "reason": "Checkpoint loading slowdown is unrelated to tensor-parallel training bugs." + }, + { + "left": "issue:44230", + "right": "issue:45310", + "accept": false, + "reason": "FP8 inference support and Qwen3.5 MoE from_pretrained failure are different model features/issues." + }, + { + "left": "issue:43298", + "right": "issue:44230", + "accept": false, + "reason": "Continuous batching for audio models is unrelated to Qwen3-VL/Qwen3.5 FP8 support." + } + ] + }, + "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": 20, + "review_unit_index": 54, + "review_unit_count": 54, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-10000-459", + "nodes": [ + "issue:37934", + "issue:42936", + "issue:43240", + "issue:43668", + "issue:43717", + "issue:43813", + "issue:44031", + "issue:46018", + "issue:46144" + ], + "soft_pairs": [ + "issue:43717|issue:46144", + "issue:42936|issue:46018", + "issue:43668|issue:44031", + "issue:42936|issue:43240", + "issue:43240|issue:46144", + "issue:42936|issue:44031", + "issue:42936|issue:46144", + "issue:42936|issue:43813", + "issue:43240|issue:43668", + "issue:43668|issue:43813", + "issue:37934|issue:43668", + "issue:42936|issue:43668" + ], + "prepared_review_unit_hash": "7f8c2d8e9e743f6bd8b9b4dd7f41af4b3b975f92d4731fb03800125def60fd79", + "component_budget": { + "node_count": 459, + "item_count": 459, + "soft_pair_count": 1688, + "serialized_chars": 875247, + "estimated_input_tokens": 218812, + "estimated_eval_tokens": 437880 + }, + "budget": { + "node_count": 9, + "item_count": 9, + "soft_pair_count": 12, + "serialized_chars": 8610, + "estimated_input_tokens": 2153, + "estimated_eval_tokens": 4562 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "The cluster is a loose mix of unrelated model, config, tokenizer, and utility issues. None of the soft pairs look like the same underlying bug or change, so no duplicates should be merged here.", + "confidence": 0.93, + "canonical_issue_reason": "Issue 44031 is the clearest standalone bug report: it describes a concrete, user-facing regression with a specific version trigger and broad tokenizer impact.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 44031 is the strongest representative of the set because it is specific, actionable, and broadly impactful compared with the more model-local or narrow reports.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43717", + "right": "issue:46144", + "accept": false, + "reason": "Different problems in different model components: init-weights behavior for Mamba/Mamba-2 vs RoFormer attention not using the attention interface." + }, + { + "left": "issue:42936", + "right": "issue:46018", + "accept": false, + "reason": "Mask2former ignore_value handling and DeepSeek-V4 shared-expert gating are unrelated code paths and symptoms." + }, + { + "left": "issue:43668", + "right": "issue:44031", + "accept": false, + "reason": "A config type-hint issue in ModernBERT is not the same as a tokenizer regex warning regression affecting all tokenizers." + }, + { + "left": "issue:42936", + "right": "issue:43240", + "accept": false, + "reason": "Both mention loss-related plumbing, but one is Mask2former ignore_value usage and the other is generic kwargs propagation to a loss function." + }, + { + "left": "issue:43240", + "right": "issue:46144", + "accept": false, + "reason": "Loss-function kwargs forwarding and RoFormer attention-interface usage are different subsystems and failure modes." + }, + { + "left": "issue:42936", + "right": "issue:44031", + "accept": false, + "reason": "Mask2former loss-argument handling is unrelated to a tokenizer regex warning after version 4.57.3." + }, + { + "left": "issue:42936", + "right": "issue:46144", + "accept": false, + "reason": "Unrelated model areas: Mask2former loss handling vs RoFormer attention implementation." + }, + { + "left": "issue:42936", + "right": "issue:43813", + "accept": false, + "reason": "Mask2former ignore_value usage and a quantization-attribute typo are not the same underlying bug." + }, + { + "left": "issue:43240", + "right": "issue:43668", + "accept": false, + "reason": "Generic loss kwargs propagation and a ModernBERT type-hint correction are distinct, with no shared code path." + }, + { + "left": "issue:43668", + "right": "issue:43813", + "accept": false, + "reason": "A type-hint fix and a typo in a quantization symbol are different kinds of issues in different areas." + }, + { + "left": "issue:37934", + "right": "issue:43668", + "accept": false, + "reason": "Both touch normalization-related terminology, but one questions Llama4TextL2Norm semantics while the other is just a ModernBERT config annotation bug." + }, + { + "left": "issue:42936", + "right": "issue:43668", + "accept": false, + "reason": "Mask2former ignore_value logic and ModernBERT norm_eps typing are unrelated model bugs." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 2, + "component_count": 20, + "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": "a78dc1631eeeb3b190fc7ba626564ae69140d9ea1ad1bad2e6d62eb3f7b6e2ba", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 2876, + "estimated_input_tokens": 719, + "estimated_eval_tokens": 1694 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The cluster does not contain duplicates: all three issues ask for different features in different parts of the library.", + "confidence": 0.97, + "canonical_issue_reason": "No canonical issue fits well because the issues are unrelated: tokenizer truncation behavior, ReduceLROnPlateau-like scheduling, and Wav2Vec2 word time offsets.", + "canonical_pr_reason": null, + "best_issue_reason": "None of the issues is a strong cluster representative since they do not describe the same underlying bug or change.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:10519", + "right": "issue:11307", + "accept": false, + "reason": "Different subsystems and requests: truncation-from-beginning for tokenization vs word boundary timestamps for Wav2Vec2." + }, + { + "left": "issue:10519", + "right": "issue:11005", + "accept": false, + "reason": "Unrelated functionality: tokenizer truncation options vs a ReduceLROnPlateau-like scheduler feature." + }, + { + "left": "issue:11005", + "right": "issue:11307", + "accept": false, + "reason": "Separate feature requests in different model areas with no shared concrete code-path or bug." + } + ] + }, + "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": 20, + "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": "688441b0425578ffed0753a2c40d4511b5e6ead1097a29d8bd154b018ab6deb0", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 2478, + "estimated_input_tokens": 620, + "estimated_eval_tokens": 1496 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These three issues do not appear to be duplicates of one another: one is about Tapas tokenization mismatch, one about a private-model download/auth error, and one about failing to load bert-base-uncased. The only commonality is generic model-loading context, not the same underlying bug.", + "confidence": 0.96, + "canonical_issue_reason": "If forced to choose a representative issue, #16618 is the broadest and most actively discussed model-loading report, but it still does not subsume the other two topics.", + "canonical_pr_reason": null, + "best_issue_reason": "#16618 is the best single issue to keep as the cluster representative because it is the most generic and most discussed of the three, though it should not be treated as a true duplicate target for the others.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:13244", + "right": "issue:16618", + "accept": false, + "reason": "Tapas tokenization behavior is a different subsystem and bug from a generic bert-base-uncased loading failure; no shared code path is evident." + }, + { + "left": "issue:15980", + "right": "issue:16618", + "accept": false, + "reason": "The private-model unauthorized download error is an auth/error-messaging problem, while #16618 is a model loading failure for a public model; these do not look like the same bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 4, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-26350-2", + "nodes": [ + "issue:26350", + "issue:33260" + ], + "soft_pairs": [ + "issue:26350|issue:33260" + ], + "prepared_review_unit_hash": "2c4f02c24ae771a174e046bec7589c8c8d77fe3bc3f3b475b0536e09e9dd5773", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1590, + "estimated_input_tokens": 398, + "estimated_eval_tokens": 1052 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two separate community contribution issues about adding support for different technologies: Flash Attention 2 vs GGUF. They share a template-like title but not the same underlying bug or change request, so they should not be clustered as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": "Neither issue is a duplicate of the other; they request different support additions for different features, so there is no clear canonical issue within this pair.", + "canonical_pr_reason": null, + "best_issue_reason": "If one must be chosen for the cluster, #26350 is slightly older and more broadly discussed, but it is not a better duplicate representative because the issues are unrelated.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:26350", + "right": "issue:33260", + "accept": false, + "reason": "Same issue template, but different feature areas and different requested changes; not the same underlying bug or enhancement." + } + ] + }, + "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": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-27088-8", + "nodes": [ + "issue:27088", + "issue:28936", + "issue:38435", + "issue:43834", + "issue:44365", + "issue:44370", + "issue:46222", + "issue:46226" + ], + "soft_pairs": [ + "issue:27088|issue:44370", + "issue:27088|issue:44365", + "issue:27088|issue:43834", + "issue:27088|issue:38435", + "issue:27088|issue:28936", + "issue:28936|issue:43834", + "issue:28936|issue:44365", + "issue:28936|issue:38435", + "issue:28936|issue:44370", + "issue:38435|issue:44370", + "issue:38435|issue:44365", + "issue:38435|issue:43834", + "issue:38435|issue:46222", + "issue:38435|issue:46226", + "issue:43834|issue:44370", + "issue:43834|issue:44365", + "issue:43834|issue:46222", + "issue:43834|issue:46226", + "issue:44365|issue:44370", + "issue:44365|issue:46222", + "issue:44365|issue:46226", + "issue:44370|issue:46222", + "issue:44370|issue:46226", + "issue:46222|issue:46226" + ], + "prepared_review_unit_hash": "4700cac3a543216ab8e79ab1e4b6f909bb14ce3f931852f8a89103b9c5cad8a6", + "component_budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13525, + "estimated_input_tokens": 3382, + "estimated_eval_tokens": 7020 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 24, + "serialized_chars": 13525, + "estimated_input_tokens": 3382, + "estimated_eval_tokens": 7020 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster mixes several distinct i18n translation requests (Turkish, Spanish, Romanian) with two separate exact-duplicate subgroups: the placeholder-title translation issues and the repeated Thai-titled issue. Only the exact matches should be merged.", + "confidence": 0.88, + "canonical_issue_reason": "Issue 43834 is the earliest of the exact-duplicate placeholder-title translation requests, making it the best canonical representative for that duplicate subgroup.", + "canonical_pr_reason": null, + "best_issue_reason": "43834 is the most suitable issue to keep because it is the earliest, generic representative of the duplicated placeholder-title translation request set.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:27088", + "right": "issue:44370", + "accept": false, + "reason": "Turkish translation request vs. generic placeholder-title issue; no evidence they describe the same underlying request." + }, + { + "left": "issue:27088", + "right": "issue:44365", + "accept": false, + "reason": "Turkish translation request vs. generic placeholder-title issue; only broad i18n similarity." + }, + { + "left": "issue:27088", + "right": "issue:43834", + "accept": false, + "reason": "Concrete Turkish docs translation request is not the same as the placeholder-title issue." + }, + { + "left": "issue:27088", + "right": "issue:38435", + "accept": false, + "reason": "Different language requests (Turkish vs Romanian); not the same issue." + }, + { + "left": "issue:27088", + "right": "issue:28936", + "accept": false, + "reason": "Different language requests (Turkish vs Spanish); not duplicates." + }, + { + "left": "issue:28936", + "right": "issue:43834", + "accept": false, + "reason": "Spanish translation request vs. generic placeholder-title issue; insufficient evidence of same underlying issue." + }, + { + "left": "issue:28936", + "right": "issue:44365", + "accept": false, + "reason": "Spanish translation request vs. generic placeholder-title issue; only topic overlap." + }, + { + "left": "issue:28936", + "right": "issue:38435", + "accept": false, + "reason": "Different language requests (Spanish vs Romanian); not the same underlying issue." + }, + { + "left": "issue:28936", + "right": "issue:44370", + "accept": false, + "reason": "Spanish translation request vs. generic placeholder-title issue; not a concrete duplicate match." + }, + { + "left": "issue:38435", + "right": "issue:44370", + "accept": false, + "reason": "Romanian translation request vs. placeholder-title issue; no direct duplication evidence." + }, + { + "left": "issue:38435", + "right": "issue:44365", + "accept": false, + "reason": "Romanian translation request vs. placeholder-title issue; not the same issue." + }, + { + "left": "issue:38435", + "right": "issue:43834", + "accept": false, + "reason": "Romanian translation request vs. placeholder-title issue; only broad i18n overlap." + }, + { + "left": "issue:38435", + "right": "issue:46222", + "accept": false, + "reason": "Romanian docs translation request and Thai-titled issue are unrelated." + }, + { + "left": "issue:38435", + "right": "issue:46226", + "accept": false, + "reason": "Romanian docs translation request and Thai-titled issue are unrelated." + }, + { + "left": "issue:43834", + "right": "issue:44370", + "accept": true, + "reason": "Exact same placeholder title and near-identical body shape; likely duplicate issues for the same translation template." + }, + { + "left": "issue:43834", + "right": "issue:44365", + "accept": true, + "reason": "Same placeholder title/body pattern; looks like a duplicate of the same translation template issue." + }, + { + "left": "issue:43834", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; different content and no same-request evidence." + }, + { + "left": "issue:43834", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not the same underlying issue." + }, + { + "left": "issue:44365", + "right": "issue:44370", + "accept": true, + "reason": "Exact same placeholder title and highly similar bodies; strong duplicate signal." + }, + { + "left": "issue:44365", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; no basis to merge." + }, + { + "left": "issue:44365", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not duplicates." + }, + { + "left": "issue:44370", + "right": "issue:46222", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; unrelated request text." + }, + { + "left": "issue:44370", + "right": "issue:46226", + "accept": false, + "reason": "Placeholder-title translation issue vs. Thai-titled issue; not the same issue." + }, + { + "left": "issue:46222", + "right": "issue:46226", + "accept": true, + "reason": "Identical Thai title and matching body size/activity; this looks like a direct duplicate." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The cluster summary is grounded in the packet: the three placeholder-title issues form a likely duplicate subgroup with 43834 as the earliest representative, and 46222/46226 are an exact-title duplicate pair. The Turkish, Spanish, and Romanian issues are kept separate, and the rejected cross-language/cross-template pairs are conservatively argued. 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": 6, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-36817-16", + "nodes": [ + "issue:36817", + "issue:36879", + "issue:37015", + "issue:41115", + "issue:41211", + "issue:42503", + "issue:42738", + "issue:43671", + "issue:43837", + "issue:43909", + "issue:44121", + "issue:44205", + "issue:45027", + "issue:45177", + "issue:45322", + "issue:46133" + ], + "soft_pairs": [ + "issue:36817|issue:41211", + "issue:36817|issue:43671", + "issue:36817|issue:36879", + "issue:36879|issue:41211", + "issue:36879|issue:43909", + "issue:36879|issue:44121", + "issue:36879|issue:43837", + "issue:36879|issue:43671", + "issue:37015|issue:43837", + "issue:37015|issue:44121", + "issue:37015|issue:43909", + "issue:36879|issue:37015", + "issue:37015|issue:42503", + "issue:41115|issue:43837", + "issue:41115|issue:44121", + "issue:41115|issue:42503", + "issue:41115|issue:45027", + "issue:41211|issue:43671", + "issue:42503|issue:43837", + "issue:42503|issue:43671", + "issue:42503|issue:42738", + "issue:42738|issue:45027", + "issue:43671|issue:43837", + "issue:43671|issue:45027", + "issue:43671|issue:45177", + "issue:43671|issue:44121", + "issue:43837|issue:45322", + "issue:43837|issue:43909", + "issue:43909|issue:44205", + "issue:43671|issue:43909", + "issue:44121|issue:45322", + "issue:43837|issue:44121", + "issue:43837|issue:45027", + "issue:37015|issue:45027", + "issue:43837|issue:45177", + "issue:45322|issue:46133" + ], + "prepared_review_unit_hash": "ab5b1787b5d2f98cd3c22a9cff89910dc09a82a92f7e2a7ea09f9a7ba7ff821b", + "component_budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21654, + "estimated_input_tokens": 5414, + "estimated_eval_tokens": 11084 + }, + "budget": { + "node_count": 16, + "item_count": 16, + "soft_pair_count": 36, + "serialized_chars": 21654, + "estimated_input_tokens": 5414, + "estimated_eval_tokens": 11084 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are all broad model-addition requests, but they target different concrete models, modalities, or architecture changes. They are related thematically, not duplicates, so no soft pair should be merged. No PRs are present.", + "confidence": 0.94, + "canonical_issue_reason": "issue:36879 is a clean, concrete representative of the cluster\u2019s general \u2018add model support\u2019 theme, but it does not subsume the others.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:36879 is the best representative anchor: short, specific, and easy to recognize as a model-support request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:36817", + "right": "issue:41211", + "accept": false, + "reason": "Different model requests: EuroBert vs DEIMv2." + }, + { + "left": "issue:36817", + "right": "issue:43671", + "accept": false, + "reason": "Different targets and modalities: EuroBert vs Qwen3-TTS." + }, + { + "left": "issue:36817", + "right": "issue:36879", + "accept": false, + "reason": "Different models: EuroBert vs RF-DETR." + }, + { + "left": "issue:36879", + "right": "issue:41211", + "accept": false, + "reason": "Different detector/model requests: RF-DETR vs DEIMv2." + }, + { + "left": "issue:36879", + "right": "issue:43909", + "accept": false, + "reason": "Different modality and code path: RF-DETR vs LFM2.5 Audio." + }, + { + "left": "issue:36879", + "right": "issue:44121", + "accept": false, + "reason": "Different change type: model support vs sparse-transformer architecture request." + }, + { + "left": "issue:36879", + "right": "issue:43837", + "accept": false, + "reason": "Different modalities: RF-DETR vs Qwen3-ASR." + }, + { + "left": "issue:36879", + "right": "issue:43671", + "accept": false, + "reason": "Different modalities: RF-DETR vs Qwen3-TTS." + }, + { + "left": "issue:37015", + "right": "issue:43837", + "accept": false, + "reason": "Different requests: NeoBERT vs Qwen3-ASR." + }, + { + "left": "issue:37015", + "right": "issue:44121", + "accept": false, + "reason": "Different underlying changes: NeoBERT support vs sparse-transformer support." + }, + { + "left": "issue:37015", + "right": "issue:43909", + "accept": false, + "reason": "Different domains: NeoBERT vs audio-model support." + }, + { + "left": "issue:37015", + "right": "issue:42503", + "accept": false, + "reason": "Both are BERT-like, but they request distinct models and likely distinct integration work." + }, + { + "left": "issue:41115", + "right": "issue:43837", + "accept": false, + "reason": "Different modalities: MiniCPM3 architecture vs ASR support." + }, + { + "left": "issue:41115", + "right": "issue:44121", + "accept": false, + "reason": "Different model families and changes: MiniCPM3 vs sparse transformer." + }, + { + "left": "issue:41115", + "right": "issue:42503", + "accept": false, + "reason": "Different model requests: MiniCPM3 vs ModernVBERT." + }, + { + "left": "issue:41115", + "right": "issue:45027", + "accept": false, + "reason": "Different modalities: MiniCPM3 vs Voxtral TTS." + }, + { + "left": "issue:41211", + "right": "issue:43671", + "accept": false, + "reason": "Different model families and tasks: DEIMv2 vs Qwen3-TTS." + }, + { + "left": "issue:42503", + "right": "issue:43837", + "accept": false, + "reason": "Different tasks/modalities: ModernVBERT vs Qwen3-ASR." + }, + { + "left": "issue:42503", + "right": "issue:43671", + "accept": false, + "reason": "Different tasks/modalities: ModernVBERT vs Qwen3-TTS." + }, + { + "left": "issue:42503", + "right": "issue:42738", + "accept": false, + "reason": "Related BERT theme, but one is a generic RoPE request and the other is a specific model request." + }, + { + "left": "issue:42738", + "right": "issue:45027", + "accept": false, + "reason": "Unrelated requests: generic BERT-like RoPE support vs Voxtral TTS." + }, + { + "left": "issue:43671", + "right": "issue:43837", + "accept": false, + "reason": "Related audio family, but TTS and ASR are distinct model/code paths." + }, + { + "left": "issue:43671", + "right": "issue:45027", + "accept": false, + "reason": "Both are TTS-related, but they are different models and integration work." + }, + { + "left": "issue:43671", + "right": "issue:45177", + "accept": false, + "reason": "Different modalities: TTS support vs OCR2 support." + }, + { + "left": "issue:43671", + "right": "issue:44121", + "accept": false, + "reason": "Different change types: TTS model support vs sparse-transformer model request." + }, + { + "left": "issue:43837", + "right": "issue:45322", + "accept": false, + "reason": "Different modalities: ASR support vs perception-encoder support." + }, + { + "left": "issue:43837", + "right": "issue:43909", + "accept": false, + "reason": "Related audio area, but distinct models and integration paths." + }, + { + "left": "issue:43909", + "right": "issue:44205", + "accept": false, + "reason": "Different domains: audio model support vs SAM3-LiteText." + }, + { + "left": "issue:43671", + "right": "issue:43909", + "accept": false, + "reason": "Different audio-model requests: Qwen3-TTS vs LFM2.5 Audio." + }, + { + "left": "issue:44121", + "right": "issue:45322", + "accept": false, + "reason": "Unrelated model families: sparse transformer vs perception encoder." + }, + { + "left": "issue:43837", + "right": "issue:44121", + "accept": false, + "reason": "Different requests: ASR support vs sparse-transformer support." + }, + { + "left": "issue:43837", + "right": "issue:45027", + "accept": false, + "reason": "Both audio-related, but ASR and TTS are distinct changes." + }, + { + "left": "issue:37015", + "right": "issue:45027", + "accept": false, + "reason": "Different domains: NeoBERT support vs Voxtral TTS support." + }, + { + "left": "issue:43837", + "right": "issue:45177", + "accept": false, + "reason": "Different tasks/modalities: ASR support vs OCR2 support." + }, + { + "left": "issue:45322", + "right": "issue:46133", + "accept": false, + "reason": "Different vision-language model requests: EUPE vs TIPSv2." + } + ] + }, + "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": 20, + "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": "8e212551ab9be4ffb6841a92be4b0511521a70b31d57eb0458e8246c631581d4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1505, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1505, + "estimated_input_tokens": 377, + "estimated_eval_tokens": 1010 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are not duplicates: one is a docs request to add Pruna as an optimization framework, while the other is a Vision RoPE refactor. They do not describe the same underlying bug or change.", + "confidence": 0.99, + "canonical_issue_reason": "No canonical issue: the two issues are unrelated in scope and purpose (documentation request vs model internals refactor).", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a good canonical duplicate target because they concern different subsystems and different kinds of work.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:38740", + "right": "issue:46443", + "accept": false, + "reason": "Different topics and change types; the first is a docs/optimization-framework request, the second is a Vision RoPE code refactor." + } + ] + }, + "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": 20, + "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": "d8f480ddb7b81dad3662f6b227689f6f78ad257ff090d169f89813905215ffc7", + "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 a generic \"Set Block Decoding\" request, while the other is specifically about adding DeepSeek v3.2 support. They describe different user needs and likely different code paths.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": "Issue #43037 is the more concrete and actionable request; #41084 is too vague to serve as a strong canonical representative.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:41084", + "right": "issue:43037", + "accept": false, + "reason": "Different underlying topics: block decoding behavior vs model/version support. No clear shared bug or change scope." + } + ] + }, + "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": 20, + "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": "909006205d4b5a0aa0ed6950914cf26847e07222a740124f4a84b1005146d76a", + "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 related to Trainer customization, but they describe different problems: one about `remove_unused_column` behavior, the other about FLOPS estimation for custom trainers. They should not be deduplicated.", + "confidence": 0.91, + "canonical_issue_reason": "Issue #43598 is the broader, more central Trainer customizability request; #43600 is a separate FLOPS-estimation concern rather than the same bug.", + "canonical_pr_reason": null, + "best_issue_reason": "#43598 is the better canonical issue because it frames the main Trainer customization limitation more generally.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43598", + "right": "issue:43600", + "accept": false, + "reason": "Both mention custom trainers/Trainer internals, but one is about `remove_unused_column` customizability and the other is about estimating FLOPS. Different code paths and different fixes." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 10, + "component_count": 20, + "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": "5fa4b56f9e4eb15d356ef9cb5dec3bd0f7f7fba3a8d0e5db7dbfb4e058c9ec33", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1612, + "estimated_input_tokens": 403, + "estimated_eval_tokens": 1062 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These issues are related only at a very broad level (adding AutoModel support for Qwen variants), but they target different model families and different auto-class behaviors, so they are not duplicates.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #43684 is the broader, more foundational registration request (AutoModel and AutoModelForConditionalGeneration for Qwen3-Omni) and is the better canonical issue if one must be chosen.", + "canonical_pr_reason": null, + "best_issue_reason": "#43684 is the most central issue in this pair because it asks for core model registration rather than a narrower task-specific classifier mapping.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43684", + "right": "issue:44405", + "accept": false, + "reason": "Different underlying request: Qwen3-Omni model registration for AutoModel/AutoModelForConditionalGeneration vs. Qwen3.5 sequence-classification support for AutoModelForSequenceClassification. Different model family and 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": 11, + "component_count": 20, + "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": "11710777dcc152416c6a73c67cdc30a73e2ac49e49e39bbf858258c03ea27f21", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 4267, + "estimated_input_tokens": 1067, + "estimated_eval_tokens": 2390 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "One low-signal cluster: three issues share the exact title \"huggingface\" and look related by repetition, while \"Fett\" appears unrelated. Evidence is too sparse to confidently merge any pair.", + "confidence": 0.41, + "canonical_issue_reason": "Issue 43845 is the earliest of the repeated \"huggingface\" issues and is the most reasonable representative of that subgroup; 43835 has a different title and stands apart.", + "canonical_pr_reason": null, + "best_issue_reason": "43845 is the best issue to keep as the cluster representative because it is the earliest identical-title report and there is no richer content to distinguish the duplicates.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:43835", + "right": "issue:43845", + "accept": false, + "reason": "Different titles and no substantive body detail tying \"Fett\" to the repeated \"huggingface\" reports." + }, + { + "left": "issue:43835", + "right": "issue:43846", + "accept": false, + "reason": "Different titles and no evidence they describe the same underlying problem." + }, + { + "left": "issue:43835", + "right": "issue:43859", + "accept": false, + "reason": "No meaningful overlap beyond being closed issues; the titles do not suggest the same bug." + }, + { + "left": "issue:43845", + "right": "issue:43846", + "accept": false, + "reason": "These look duplicate-like because the titles match, but there is no issue description detail to confirm the same underlying bug or change." + }, + { + "left": "issue:43845", + "right": "issue:43859", + "accept": false, + "reason": "Exact same title, but the reports are too content-poor to verify they refer to the same concrete issue." + }, + { + "left": "issue:43846", + "right": "issue:43859", + "accept": false, + "reason": "Exact same title, yet the available evidence is insufficient to assert a shared 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": 12, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44371-5", + "nodes": [ + "issue:44371", + "issue:44418", + "issue:44572", + "issue:44573", + "issue:44937" + ], + "soft_pairs": [ + "issue:44371|issue:44572", + "issue:44371|issue:44573", + "issue:44371|issue:44418", + "issue:44371|issue:44937", + "issue:44418|issue:44572", + "issue:44418|issue:44573", + "issue:44572|issue:44573", + "issue:44572|issue:44937", + "issue:44573|issue:44937" + ], + "prepared_review_unit_hash": "fe89be851ada82bfb850dc24de646bfcc731f36b386f1ecc855ec9cde2e19115", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 9, + "serialized_chars": 5788, + "estimated_input_tokens": 1447, + "estimated_eval_tokens": 3150 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly spam/placeholder issues with no evidence of shared bug reports. Issue 44418 is the only one with a meaningful title, so it is the best anchor, but none of the pairings look like true duplicates.", + "confidence": 0.94, + "canonical_issue_reason": "issue:44418 is the only non-spam, descriptive issue in the set, so it is the best canonical anchor among otherwise low-signal reports.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:44418 is the most substantive and least noisy issue here; the others are generic spam or similarly uninformative.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:44371", + "right": "issue:44572", + "accept": false, + "reason": "Both are generic '' placeholders with no concrete shared bug/change to deduplicate." + }, + { + "left": "issue:44371", + "right": "issue:44573", + "accept": false, + "reason": "Both are generic '' placeholders with no concrete shared bug/change to deduplicate." + }, + { + "left": "issue:44371", + "right": "issue:44418", + "accept": false, + "reason": "One is spam, the other is a documentation suggestion; the titles do not indicate the same underlying issue." + }, + { + "left": "issue:44371", + "right": "issue:44937", + "accept": false, + "reason": "Generic spam is not the same underlying issue as the 'Google LLC' title, which is a different topic entirely." + }, + { + "left": "issue:44418", + "right": "issue:44572", + "accept": false, + "reason": "A documentation enhancement request does not match a generic spam placeholder." + }, + { + "left": "issue:44418", + "right": "issue:44573", + "accept": false, + "reason": "A documentation enhancement request does not match a generic spam placeholder." + }, + { + "left": "issue:44572", + "right": "issue:44573", + "accept": false, + "reason": "Both are spam placeholders, but there is no evidence they describe the same concrete bug or change." + }, + { + "left": "issue:44572", + "right": "issue:44937", + "accept": false, + "reason": "Generic spam and a 'Google LLC' topic are unrelated on the available evidence." + }, + { + "left": "issue:44573", + "right": "issue:44937", + "accept": false, + "reason": "Generic spam and a 'Google LLC' topic are unrelated on the available evidence." + } + ] + }, + "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": 20, + "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": "a1817adf62117243a2d32e5f94bf25a32fe2c69ea6edf7548c18fc59c763e87f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1645, + "estimated_input_tokens": 412, + "estimated_eval_tokens": 1080 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues look highly similar, but the packet only supports a cautious association based on identical titles and soft-similarity signals. #45145 is the better representative because it is earlier and has the only inbound reference, but the evidence is not strong enough to confidently assert a duplicate relationship.", + "confidence": 0.74, + "canonical_issue_reason": "#45145 is the earlier issue and has the only inbound reference, so it is the best representative of the pair if they are ultimately treated as related. The packet does not include bodies, so this should be framed as a representative choice rather than a proven duplicate determination.", + "canonical_pr_reason": null, + "best_issue_reason": "#45145 is the best issue representative because it is earlier and has an inbound reference, while the available evidence does not show any substantive difference from #45151. This is a conservative selection rather than a strong duplicate claim.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45145", + "right": "issue:45151", + "accept": false, + "reason": "The identical title and soft-similarity signal suggest a possible duplicate, but the packet does not provide issue bodies and the pair is marked deterministic_accept:false. That is not enough evidence to confidently accept the edge as the same underlying issue." + } + ] + }, + "evaluator_result": { + "accept": false, + "feedback": "The verdict is too strong for the evidence in the packet. We only have same title, similar body lengths, and a soft-similarity signal; the packet does not provide the issue bodies, so claims like \"near-identical content\" or \"clear duplicates\" are not fully grounded. Also, the soft pair is marked `deterministic_accept: false`, so accepting the duplicate edge is not conservative enough. The canonical-choice rationale can mention that #45145 is earlier and has an inbound reference, but it should not overstate substantiated duplication." + }, + "evaluator_used": true, + "retried": true, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 14, + "component_count": 20, + "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": 15, + "component_count": 20, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45672-2", + "nodes": [ + "issue:45672", + "issue:45674" + ], + "soft_pairs": [ + "issue:45672|issue:45674" + ], + "prepared_review_unit_hash": "0344bfe59ff04a353e88848e8f983e436f5fa7ad0ace78e7abca33021bda56ab", + "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 are two separate 4-bit inference bugs: one about torch.finfo() failing on uint8 weights in Gemma4 audio modules, the other about llm_int8_skip_modules wiping the default lm_head exclusion and triggering an AssertionError. They are related only at a broad BitsAndBytes/4-bit level, not the same underlying defect.", + "confidence": 0.94, + "canonical_issue_reason": "Issue #45674 is the better representative because it describes a concrete configuration regression in BitsAndBytesConfig with a clearer failure mode and broader impact.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue #45674 is the strongest standalone issue in the set: it is specific, actionable, and maps to a distinct code-path problem rather than a model-specific symptom.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45672", + "right": "issue:45674", + "accept": false, + "reason": "Reject: the failures are different. #45672 is a Gemma4/audio-module torch.finfo() uint8-type issue, while #45674 is an llm_int8_skip_modules / lm_head exclusion regression causing AssertionError. Same subsystem, different bug." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 16, + "component_count": 20, + "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": "5ee06a725d5ffa0809c5e835c2dc67cb5cc297018db98101d007d69e63b52ed3", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1477, + "estimated_input_tokens": 370, + "estimated_eval_tokens": 996 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two closed issues with unrelated personal-name titles and different dates; they do not describe the same bug or request.", + "confidence": 0.98, + "canonical_issue_reason": "Issue 45762 is a reasonable canonical only by default; it has no body or technical context, so there is no substantive basis to prefer it over 45763.", + "canonical_pr_reason": null, + "best_issue_reason": "Neither issue is a strong canonical bug report, but 45762 is marginally as good a representative as 45763 because both are empty, non-technical, and equally non-descriptive.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45762", + "right": "issue:45763", + "accept": false, + "reason": "The titles are different personal names with different dates and no shared technical details, explicit links, or overlapping behavior; they appear unrelated rather than duplicate." + } + ] + }, + "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": 20, + "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": "74fb955d415de602c7ee447936971f0274f873516dd94ae2e8b7d45a3db29880", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1481, + "estimated_input_tokens": 371, + "estimated_eval_tokens": 998 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two closed issues are just name/date placeholders and do not describe a shared technical problem, so they should not be clustered as duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45764", + "right": "issue:45765", + "accept": false, + "reason": "No bug report, feature request, or code-path description is present; the titles only contain different personal names and dates, so there is no evidence they refer to 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": 20, + "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": "e7ed51b01c7d79a02b2d10d3b7410eec7d9196461c2ca1ccde215c243b887b31", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1540, + "estimated_input_tokens": 385, + "estimated_eval_tokens": 1026 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are not duplicates: one requests a new Qwen3_5 token-classification/value-model class, while the other asks for FP8 training support. They touch different features and code paths.", + "confidence": 0.95, + "canonical_issue_reason": "Issue #46295 is the broader, more general feature request; #45810 is a narrow model-specific addition and does not match it.", + "canonical_pr_reason": null, + "best_issue_reason": "#46295 is the better representative issue for the cluster because it is the more general and potentially higher-impact request.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45810", + "right": "issue:46295", + "accept": false, + "reason": "Different underlying requests: adding a Qwen3_5 token-classification/value-model class versus enabling FP8 training. No shared bug or concrete code-path fix." + } + ] + }, + "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": 20, + "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": "344a4eabb7fbfe3230d4be74f88bd95311db751bd83bb0b5c52ef183da2b059b", + "component_budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5729, + "estimated_input_tokens": 1433, + "estimated_eval_tokens": 3122 + }, + "budget": { + "node_count": 5, + "item_count": 5, + "soft_pair_count": 8, + "serialized_chars": 5729, + "estimated_input_tokens": 1433, + "estimated_eval_tokens": 3122 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "There are two exact duplicate issue pairs: 45870/45871 and 45872/45873. Issue 45869 is thematically similar but appears to be a different dataset promotion topic, not the same underlying request/bug. No pull requests are present.", + "confidence": 0.93, + "canonical_issue_reason": "issue:45870 is the best canonical issue because it is the earliest and slightly more active of the exact duplicate pair 45870/45871; 45871 is a straight duplicate, while 45869 and 45872/45873 are separate topics.", + "canonical_pr_reason": null, + "best_issue_reason": "issue:45870 is the strongest representative issue in the cluster: it is an exact duplicate target, has a duplicate partner, and is a cleaner canonical anchor than the unrelated dataset promo issue.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:45869", + "right": "issue:45870", + "accept": false, + "reason": "Related financial/DataForge wording, but one is a dataset promotion and the other is an example request; not the same underlying issue." + }, + { + "left": "issue:45869", + "right": "issue:45871", + "accept": false, + "reason": "Same as 45869/45870: topical overlap only, not a true duplicate." + }, + { + "left": "issue:45870", + "right": "issue:45871", + "accept": true, + "reason": "Exact same title and effectively the same issue content; clear duplicate." + }, + { + "left": "issue:45870", + "right": "issue:45872", + "accept": false, + "reason": "Both are DataForge/financial-themed, but they describe different requests (example prediction vs scraping service)." + }, + { + "left": "issue:45870", + "right": "issue:45873", + "accept": false, + "reason": "Different request from 45870; similar surface wording but not the same underlying issue." + }, + { + "left": "issue:45871", + "right": "issue:45872", + "accept": false, + "reason": "Exact-duplicate partner of 45870, but not the same topic as the scraping-service issue." + }, + { + "left": "issue:45871", + "right": "issue:45873", + "accept": false, + "reason": "Different request from 45871; similar subsystem mention only." + }, + { + "left": "issue:45872", + "right": "issue:45873", + "accept": true, + "reason": "Exact same title and body shape; clear duplicate." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: 45870/45871 and 45872/45873 have identical titles and matching body lengths, while 45869 is clearly a different dataset-promo topic. The canonical choice of 45870 is conservative and supported by earlier creation time plus slightly higher discussion activity. No PRs are present. Soft-edge accepts are limited to the two strongest duplicate-like pairs and do not overreach." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "issue", + "component_index": 20, + "component_count": 20, + "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": "3c9fa5f880d5c6ad9484e7ae2d859b64e0e371e8aad335b8b994823703d04fa4", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 1637, + "estimated_input_tokens": 410, + "estimated_eval_tokens": 1076 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two issues are related by static caching and linear-attention, but they describe different failure modes: one is a shape corruption during `early_initialization`, the other is a `KeyError` for a missing `linear_attention` cache entry in Qwen3.5.", + "confidence": 0.9, + "canonical_issue_reason": "Issue 46439 is the better canonical issue because it states the broader underlying static-cache/linear-attention corruption and gives a more general bug description than the model-specific KeyError report.", + "canonical_pr_reason": null, + "best_issue_reason": "Issue 46439 is the strongest representative of the cluster: it is more general, describes the failure mechanism, and is more likely to map to the core code-path problem.", + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "issue:46439", + "right": "issue:46441", + "accept": false, + "reason": "Reject: they involve the same subsystem, but the reported failures differ materially (`RuntimeError` from wrong shapes during early initialization vs `KeyError: 'linear_attention'` in Qwen3.5 static cache), so they do not clearly share 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": 1, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-39895-14", + "nodes": [ + "pull_request:39895", + "pull_request:41855", + "pull_request:42668", + "pull_request:43251", + "pull_request:43254", + "pull_request:43424", + "pull_request:43823", + "pull_request:43995", + "pull_request:44827", + "pull_request:45170", + "pull_request:45213", + "pull_request:45655", + "pull_request:46048", + "pull_request:46267" + ], + "soft_pairs": [ + "pull_request:39895|pull_request:43424", + "pull_request:41855|pull_request:43823", + "pull_request:41855|pull_request:43424", + "pull_request:42668|pull_request:43823", + "pull_request:43251|pull_request:43823", + "pull_request:43251|pull_request:44827", + "pull_request:43424|pull_request:44827", + "pull_request:43424|pull_request:43823", + "pull_request:43823|pull_request:45655", + "pull_request:43823|pull_request:43995", + "pull_request:43995|pull_request:45213", + "pull_request:43995|pull_request:44827", + "pull_request:44827|pull_request:45170", + "pull_request:44827|pull_request:46267", + "pull_request:44827|pull_request:46048", + "pull_request:44827|pull_request:45655", + "pull_request:44827|pull_request:45213", + "pull_request:43823|pull_request:45213", + "pull_request:43251|pull_request:43254" + ], + "prepared_review_unit_hash": "c4cb72318b8deea24ca2d2844f50eba7649156d89c90a30192515568a359f5bd", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 19, + "serialized_chars": 38085, + "estimated_input_tokens": 9522, + "estimated_eval_tokens": 19300 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a grab bag of unrelated open pull requests across model additions, docs, tokenizer utilities, export tests, and loss/kernel fixes. The only true duplicate-looking pair is the two fixed_cross_entropy kwargs PRs targeting the same issue.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43251 is the clearest canonical choice: it explicitly targets issue #43240, has the more complete fix surface, and includes review activity. PR #43254 is the near-duplicate follow-up variant.", + "best_issue_reason": null, + "best_pr_reason": "PR #43251 is the best representative of the cluster\u2019s only real duplicate pair because it directly addresses the tracked bug with the more complete implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:39895", + "right": "pull_request:43424", + "accept": false, + "reason": "Both are unrelated PRs; one adds VideoPrism, the other adds an Executorch dynamic-shape test." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43823", + "accept": false, + "reason": "Tokenizer helper methods for MistralCommon and a new MobileLLM model are different changes and code paths." + }, + { + "left": "pull_request:41855", + "right": "pull_request:43424", + "accept": false, + "reason": "A tokenizer API addition and an Executorch export test are unrelated." + }, + { + "left": "pull_request:42668", + "right": "pull_request:43823", + "accept": false, + "reason": "Processor auto-loading robustness and MobileLLM model implementation are not the same underlying change." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43823", + "accept": false, + "reason": "Cross-entropy kwargs plumbing and adding MobileLLM are unrelated." + }, + { + "left": "pull_request:43251", + "right": "pull_request:44827", + "accept": false, + "reason": "The loss helper fix is unrelated to Mistral4 test/model adjustments." + }, + { + "left": "pull_request:43424", + "right": "pull_request:44827", + "accept": false, + "reason": "Executorch export tests and Mistral4 fixes do not overlap on the same bug." + }, + { + "left": "pull_request:43424", + "right": "pull_request:43823", + "accept": false, + "reason": "Executorch dynamic-shape testing and MobileLLM model addition are separate concerns." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45655", + "accept": false, + "reason": "A new model implementation and config resolution-order fix are different underlying changes." + }, + { + "left": "pull_request:43823", + "right": "pull_request:43995", + "accept": false, + "reason": "MobileLLM model addition and Falcon output-interface refactor are unrelated." + }, + { + "left": "pull_request:43995", + "right": "pull_request:45213", + "accept": false, + "reason": "Falcon refactoring and the model-creation skill/docs PR are not the same code-path change." + }, + { + "left": "pull_request:43995", + "right": "pull_request:44827", + "accept": false, + "reason": "Falcon output-capture refactor and Mistral4 test fixes are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45170", + "accept": false, + "reason": "Mistral4 test/model changes and the layrnorm->layernorm rename fix different problems." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46267", + "accept": false, + "reason": "Mistral4 test fixes and chat parsing updates are unrelated." + }, + { + "left": "pull_request:44827", + "right": "pull_request:46048", + "accept": false, + "reason": "Mistral4 test fixes and the TDT loss kernel addition are different features." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45655", + "accept": false, + "reason": "Mistral4 test changes and cls.config resolution-order fix are not the same bug." + }, + { + "left": "pull_request:44827", + "right": "pull_request:45213", + "accept": false, + "reason": "A model test PR and a model-creation skill/docs PR are unrelated." + }, + { + "left": "pull_request:43823", + "right": "pull_request:45213", + "accept": false, + "reason": "MobileLLM model addition and the model-creation skill/docs PR are not duplicates." + }, + { + "left": "pull_request:43251", + "right": "pull_request:43254", + "accept": true, + "reason": "Both PRs patch fixed_cross_entropy to forward cross_entropy kwargs for the same issue (#43240); they are duplicate implementations of the same fix." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative. The only accepted duplicate pair, PR #43251/#43254, is supported by the shared issue target (#43240) and matching loss_utils fix. The other soft-edge rejections are consistent with the file/diff contents and do not overstate duplication." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 2, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40546-3", + "nodes": [ + "pull_request:40546", + "pull_request:42781", + "pull_request:46331" + ], + "soft_pairs": [ + "pull_request:40546|pull_request:42781", + "pull_request:42781|pull_request:46331" + ], + "prepared_review_unit_hash": "82f4102e54dc3d6d535c096cc48e89a025aeb9728e1e4fb08ae42956f25e099e", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10869, + "estimated_input_tokens": 2718, + "estimated_eval_tokens": 5692 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10869, + "estimated_input_tokens": 2718, + "estimated_eval_tokens": 5692 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These PRs are not duplicates: one is the base VibeVoice implementation, one is a separate VibeVoice Realtime extension, and one is an unrelated Parakeet-RNNT change. The shared auto-registry files are broad integration overlap, not the same fix/change.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 40546 is the core VibeVoice implementation and the most established candidate to represent the VibeVoice work; the other VibeVoice PR is a separate realtime extension, and the Parakeet PR is unrelated.", + "best_issue_reason": null, + "best_pr_reason": "PR 40546 is the best representative of the cluster because it covers the main VibeVoice feature set and has the broadest, most mature implementation; the others are distinct follow-on or unrelated changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:40546", + "right": "pull_request:42781", + "accept": false, + "reason": "Related model family, but not the same underlying change: 40546 implements VibeVoice, while 42781 adds VibeVoice Realtime with additional acoustic-tokenizer and generation work. Shared registry/docs files are too broad to treat as duplicates." + }, + { + "left": "pull_request:42781", + "right": "pull_request:46331", + "accept": false, + "reason": "Different model targets and code paths: 42781 is VibeVoice Realtime, while 46331 adds Parakeet-RNNT. Overlap in auto-mapping files is incidental framework plumbing, not a duplicate fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 3, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-40755-2", + "nodes": [ + "pull_request:40755", + "pull_request:41224" + ], + "soft_pairs": [ + "pull_request:40755|pull_request:41224" + ], + "prepared_review_unit_hash": "278c89ecc870a6064c5f9f574556dd0da6db800d95ac84030b63849b941c2e20", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4676, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4676, + "estimated_input_tokens": 1169, + "estimated_eval_tokens": 2594 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two pull requests are not duplicates: one adds TimesFM covariate forecasting support, while the other adds DINOv3 image classification support. They touch different models, code paths, and user-facing features.", + "confidence": 0.99, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:40755", + "right": "pull_request:41224", + "accept": false, + "reason": "Different model families and unrelated changes: TimesFM covariate regression vs. DINOv3 image classification. They do not fix the same bug or implement the same feature." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 4, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41356-2", + "nodes": [ + "pull_request:41356", + "pull_request:44339" + ], + "soft_pairs": [ + "pull_request:41356|pull_request:44339" + ], + "prepared_review_unit_hash": "aead1308c167b0bae6ff11bf9358340baa5e76f816618383418b505a10a9cd01", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6050, + "estimated_input_tokens": 1513, + "estimated_eval_tokens": 3282 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two DEIMv2 pull requests target the same model family, but they are not the same change: PR 41356 is an early/basic model + image processor add, while PR 44339 is a much broader follow-up that adds loss code, auto-mappings, conversion tooling, modular files, and more integration work.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44339 is the more complete and representative DEIMv2 implementation; PR 41356 looks like an earlier partial version of the same model bring-up.", + "best_issue_reason": null, + "best_pr_reason": "PR 44339 is the strongest single artifact for the cluster because it subsumes the basic model work and includes the broader integration needed for a full Transformers addition.", + "soft_edge_verdicts": [ + { + "left": "pull_request:41356", + "right": "pull_request:44339", + "accept": false, + "reason": "Same model area, but not the same concrete change: 44339 expands the implementation substantially and adds new code paths beyond the basic PR 41356. They look like related iterations, not mergeable duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 5, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-41797-3", + "nodes": [ + "pull_request:41797", + "pull_request:45186", + "pull_request:45630" + ], + "soft_pairs": [ + "pull_request:41797|pull_request:45186", + "pull_request:45186|pull_request:45630" + ], + "prepared_review_unit_hash": "f06a1ded87d4ec8cef6c47b35a44ca6e0b18d293321460e6760131fb3fdbb510", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10436, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 10436, + "estimated_input_tokens": 2609, + "estimated_eval_tokens": 5474 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three separate model-addition PRs for different models (DeepSeek OCR, Isaac, Kimi2-6). The overlap is mostly boilerplate around Transformers model registration and docs, not the same underlying change, so none should be clustered as duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": null, + "soft_edge_verdicts": [ + { + "left": "pull_request:41797", + "right": "pull_request:45186", + "accept": false, + "reason": "Different model integrations with different code paths and artifacts: DeepSeek OCR adds OCR-specific model/processor files, while Isaac adds a separate vision model stack and tests. Shared auto-registration/docs boilerplate is not enough to treat them as the same PR." + }, + { + "left": "pull_request:45186", + "right": "pull_request:45630", + "accept": false, + "reason": "These add different models (Isaac vs Kimi2-6) with distinct configs, modeling code, and supporting files. The similarity is limited to generic model-registration updates, so they are not mergeable as one fix/change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 6, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42975-2", + "nodes": [ + "pull_request:42975", + "pull_request:42976" + ], + "soft_pairs": [ + "pull_request:42975|pull_request:42976" + ], + "prepared_review_unit_hash": "a0e2337b3dc977288bf39433d281c7b8fb32fddb91ae4ecf1f2a5321e3d9f34b", + "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 two separate dependency-upgrade PRs in GitHub Actions workflows: one targets Node 24 compatibility for core actions, the other updates docker/build-push-action and related workflow actions. They overlap in files but not in the underlying change, so they should not be merged as duplicates.", + "confidence": 0.91, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: #42975 and #42976 address different workflow-action upgrades with different scopes and failure modes.", + "best_issue_reason": null, + "best_pr_reason": "If one must be chosen, #42975 is the broader maintenance PR, but it still should not subsume #42976 because the actual action upgrades differ.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42975", + "right": "pull_request:42976", + "accept": false, + "reason": "They both edit GitHub Actions workflows, but the concrete changes are different: #42975 upgrades checkout/cache/upload-artifact and other core actions for Node 24 compatibility, while #42976 upgrades docker/build-push-action and a different set of workflow actions. Shared files alone do not make them the same fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 7, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-42978-4", + "nodes": [ + "pull_request:42978", + "pull_request:43448", + "pull_request:43451", + "pull_request:45490" + ], + "soft_pairs": [ + "pull_request:42978|pull_request:43448", + "pull_request:42978|pull_request:43451", + "pull_request:42978|pull_request:45490", + "pull_request:43448|pull_request:43451", + "pull_request:43448|pull_request:45490", + "pull_request:43451|pull_request:45490" + ], + "prepared_review_unit_hash": "f8ed0f1e01887f3627d90301ea393f1492c9bcfa420dac4b7e76dcf535fe2dc6", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14253, + "estimated_input_tokens": 3564, + "estimated_eval_tokens": 7384 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 14253, + "estimated_input_tokens": 3564, + "estimated_eval_tokens": 7384 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All four pull requests are separate model-addition efforts (ViT NEPA, Molmo, Molmo2, CTSM) with only superficial overlap in docs/auto-registration files. They do not look like duplicate fixes or mergeable into one PR.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "No single PR is a clear duplicate-canonical representative; these are distinct model integrations rather than the same underlying change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:42978", + "right": "pull_request:43448", + "accept": false, + "reason": "Different model additions with different architectures and filenames; shared docs/auto-mapping edits are incidental, not the same code-path change." + }, + { + "left": "pull_request:42978", + "right": "pull_request:43451", + "accept": false, + "reason": "ViT NEPA vs Molmo2 are unrelated model integrations; overlap is only in generic registry/docs plumbing." + }, + { + "left": "pull_request:42978", + "right": "pull_request:45490", + "accept": false, + "reason": "CTSM is a separate time-series model addition, not the same bug or feature as ViT NEPA." + }, + { + "left": "pull_request:43448", + "right": "pull_request:43451", + "accept": false, + "reason": "Molmo and Molmo2 are related in naming but are distinct model families with different files and implementations; not a single concrete fix." + }, + { + "left": "pull_request:43448", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo and CTSM touch broad model-registration infrastructure, but they implement different model families and code paths." + }, + { + "left": "pull_request:43451", + "right": "pull_request:45490", + "accept": false, + "reason": "Molmo2 and CTSM are unrelated model additions; shared auto/model registry updates do not indicate duplicate content." + } + ] + }, + "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": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43291-4", + "nodes": [ + "pull_request:43291", + "pull_request:43665", + "pull_request:44815", + "pull_request:46380" + ], + "soft_pairs": [ + "pull_request:43291|pull_request:43665", + "pull_request:43291|pull_request:44815", + "pull_request:43291|pull_request:46380", + "pull_request:43665|pull_request:44815", + "pull_request:43665|pull_request:46380", + "pull_request:44815|pull_request:46380" + ], + "prepared_review_unit_hash": "6f6350727c5da31780616683841dc4cd98cc9170c4e053655eb1f8b6e3169e2c", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 11142, + "estimated_input_tokens": 2786, + "estimated_eval_tokens": 5828 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 6, + "serialized_chars": 11142, + "estimated_input_tokens": 2786, + "estimated_eval_tokens": 5828 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "No pair looks like the same underlying bug/change. The similarity is mostly superficial: one PR is test-only noise, and the others target unrelated areas (Whisper tokenization, FP8 dequantization, and generation watermarking validation).", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46380 is the clearest real bug fix in the set: it addresses a specific validation/code-path issue by handling dict-form `watermarking_config` consistently in `GenerationConfig.validate()`.", + "best_issue_reason": null, + "best_pr_reason": "PR 46380 is the strongest standalone candidate overall because it fixes a concrete runtime validation problem with a narrow, understandable change. PR 44815 is also substantive, but it is broader and more multi-part.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43291", + "right": "pull_request:43665", + "accept": false, + "reason": "Unrelated changes: 43291 modifies Whisper decode behavior, while 43665 is just test assertions/noise across CLIP and ViT." + }, + { + "left": "pull_request:43291", + "right": "pull_request:44815", + "accept": false, + "reason": "Different bug domains and code paths: Whisper tokenizer decoding vs FP8/dequant loading logic." + }, + { + "left": "pull_request:43291", + "right": "pull_request:46380", + "accept": false, + "reason": "Whisper decode/tokenization changes are unrelated to generation watermarking validation." + }, + { + "left": "pull_request:43665", + "right": "pull_request:44815", + "accept": false, + "reason": "43665 is test-only and not a real implementation fix; it does not match the FP8 dequantization change in 44815." + }, + { + "left": "pull_request:43665", + "right": "pull_request:46380", + "accept": false, + "reason": "43665 is test-only noise, while 46380 fixes generation config validation." + }, + { + "left": "pull_request:44815", + "right": "pull_request:46380", + "accept": false, + "reason": "Both are real fixes, but they target different problems and code paths: FP8 dequantization/loading vs watermarking config validation." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 9, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43488-2", + "nodes": [ + "pull_request:43488", + "pull_request:43532" + ], + "soft_pairs": [ + "pull_request:43488|pull_request:43532" + ], + "prepared_review_unit_hash": "d9df4361a0b23916649243c01601edca30858e13e4b3eb620d99cf84cd8558a2", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4563, + "estimated_input_tokens": 1141, + "estimated_eval_tokens": 2538 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4563, + "estimated_input_tokens": 1141, + "estimated_eval_tokens": 2538 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs that both touch the ViT test file, but they are otherwise unrelated: one is a small setup/version/dependency formatting change, the other rewires CI/workflow and testing utilities. The shared file alone is not enough to treat them as the same change.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43532 is the more substantive and representative change in the cluster, with broader workflow/testing-utility edits beyond the shared test file.", + "best_issue_reason": null, + "best_pr_reason": "PR #43532 is the best representative PR here because it has the larger, more distinctive code-path impact and clearer scope.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43488", + "right": "pull_request:43532", + "accept": false, + "reason": "Reject: they do not fix the same concrete bug/change. #43488 is a setup/version/dependency tweak, while #43532 changes CI workflow and testing utilities; the shared ViT test file is incidental." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 10, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43612-2", + "nodes": [ + "pull_request:43612", + "pull_request:43613" + ], + "soft_pairs": [ + "pull_request:43612|pull_request:43613" + ], + "prepared_review_unit_hash": "5ca9447d0d2dbebbd91dbbeb7883ea01927b314fe655b825e4cd4df6ec4a2f01", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 6083, + "estimated_input_tokens": 1521, + "estimated_eval_tokens": 3298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two PRs share a lot of pipeline/doc scaffolding, but they implement different features: promptable concept segmentation vs promptable visual segmentation, with different model and test coverage. They are related, not duplicates.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR for duplication: 43612 targets concept segmentation, while 43613 targets visual segmentation and touches a different set of models/code paths.", + "best_issue_reason": null, + "best_pr_reason": "43613 is the better representative of the cluster because it is the broader, more complete PR and appears to define the more general visual segmentation pipeline.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43612", + "right": "pull_request:43613", + "accept": false, + "reason": "Both add similar pipeline/documentation scaffolding, but they implement different concrete features and model integrations, so they should not be merged as the same change." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 11, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43651-2", + "nodes": [ + "pull_request:43651", + "pull_request:43663" + ], + "soft_pairs": [ + "pull_request:43651|pull_request:43663" + ], + "prepared_review_unit_hash": "e3ba5f58abafd35b8e20b1472489f7e11a535f1f9a5a68d1a16c7665f92e750e", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4293, + "estimated_input_tokens": 1074, + "estimated_eval_tokens": 2404 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two distinct trainer extensibility PRs in the same file: one changes gradient-accumulation loss scaling behavior, the other exposes signature-column filtering as an override point. They do not look mergeable into one PR or like duplicates.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 43651 is the clearer standalone behavior change because it targets a concrete training-step loss-scaling path and introduces a single explicit control hook; 43663 is also valid, but it is a separate refactor/override point for data column filtering.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43651", + "right": "pull_request:43663", + "accept": false, + "reason": "Same file, but different underlying behaviors: 43651 changes gradient accumulation loss normalization, while 43663 adds a hook for signature-based column filtering. They affect separate code paths and would not plausibly be merged as one fix." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 12, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43656-2", + "nodes": [ + "pull_request:43656", + "pull_request:43842" + ], + "soft_pairs": [ + "pull_request:43656|pull_request:43842" + ], + "prepared_review_unit_hash": "48a14cf454e0a6403393ff46ed12435a2372962ac3b95088edb87ff0b3f0354f", + "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": "These PRs touch the same CLI module but appear to fix different NameError/typing problems: 43656 is a broader annotation/future-import cleanup, while 43842 specifically guards a TypeAdapter reference so pydantic is not required at runtime. They are not the same underlying change.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 43842 is the more precise representative because it targets a clearly scoped runtime NameError involving optional pydantic, whereas 43656 is a broader typing/annotation adjustment.", + "best_issue_reason": null, + "best_pr_reason": "PR 43842 is the strongest cluster representative: it has the clearest bug statement and the smallest, most targeted fix. 43656 looks like a separate typing-related change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43656", + "right": "pull_request:43842", + "accept": false, + "reason": "Same file and similar naming, but the fixes address different code-path problems: 43656 changes annotation behavior broadly, while 43842 specifically avoids a TypeAdapter runtime import/name error when pydantic is absent. They could not plausibly be merged as one duplicate PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 13, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43779-2", + "nodes": [ + "pull_request:43779", + "pull_request:43816" + ], + "soft_pairs": [ + "pull_request:43779|pull_request:43816" + ], + "prepared_review_unit_hash": "e7c674adcc031aa8883109a5dfc678b4407fb20b10764302092e384d031ba83a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4362, + "estimated_input_tokens": 1091, + "estimated_eval_tokens": 2438 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both pull requests appear to implement the same SwanLab integration change: adding run id/resume support in `integration_utils.py` for issue #43698, with one variant also expanding the docs. They are strong duplicate PR candidates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #43816 is the cleaner representative: it carries the same SwanLab id/resume support plus the accompanying documentation updates, making it the more complete version of the shared change.", + "best_issue_reason": null, + "best_pr_reason": "PR #43816 is the best global representative because it appears to include the full feature change and docs for SwanLab run id/resume support, while #43779 is the narrower implementation-focused variant.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43779", + "right": "pull_request:43816", + "accept": true, + "reason": "Same underlying SwanLab callback enhancement for id/resume support in the same file and same issue target; the doc differences do not change the core code-path change, so these are duplicate PRs." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: both PRs target issue #43698, touch the same file, and the previews indicate the same SwanLab id/resume support with #43816 adding docs. The duplicate/mergeable verdict is conservative enough given the shared target and overlapping code change." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43996", + "pull_request:43997", + "pull_request:44007", + "pull_request:44010", + "pull_request:44013", + "pull_request:44017", + "pull_request:44018", + "pull_request:44019", + "pull_request:44044", + "pull_request:44056", + "pull_request:44059", + "pull_request:44066", + "pull_request:44068", + "pull_request:44071", + "pull_request:44072", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44085", + "pull_request:44086", + "pull_request:44098", + "pull_request:44101", + "pull_request:44116", + "pull_request:44129", + "pull_request:44154", + "pull_request:44161", + "pull_request:44722" + ], + "soft_pairs": [ + "pull_request:43996|pull_request:44085", + "pull_request:43996|pull_request:44044", + "pull_request:44066|pull_request:44085", + "pull_request:44007|pull_request:44072", + "pull_request:44072|pull_request:44722", + "pull_request:44066|pull_request:44072", + "pull_request:44013|pull_request:44044", + "pull_request:44066|pull_request:44086", + "pull_request:44066|pull_request:44071", + "pull_request:44018|pull_request:44068", + "pull_request:44066|pull_request:44068", + "pull_request:44018|pull_request:44066", + "pull_request:44019|pull_request:44722", + "pull_request:44086|pull_request:44722", + "pull_request:44019|pull_request:44085", + "pull_request:44018|pull_request:44019", + "pull_request:44019|pull_request:44071", + "pull_request:44068|pull_request:44116", + "pull_request:44068|pull_request:44722", + "pull_request:44018|pull_request:44116", + "pull_request:44071|pull_request:44722", + "pull_request:44018|pull_request:44085", + "pull_request:44018|pull_request:44086", + "pull_request:44068|pull_request:44086", + "pull_request:44071|pull_request:44085", + "pull_request:44059|pull_request:44086", + "pull_request:44085|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:44161", + "pull_request:44017|pull_request:44722", + "pull_request:44017|pull_request:44073", + "pull_request:44068|pull_request:44071", + "pull_request:43997|pull_request:44073", + "pull_request:44071|pull_request:44086", + "pull_request:44073|pull_request:44101", + "pull_request:44017|pull_request:44071", + "pull_request:44059|pull_request:44071", + "pull_request:44073|pull_request:44161", + "pull_request:44056|pull_request:44101", + "pull_request:44018|pull_request:44059", + "pull_request:44059|pull_request:44068", + "pull_request:44101|pull_request:44161", + "pull_request:43997|pull_request:44017", + "pull_request:43997|pull_request:44056", + "pull_request:44071|pull_request:44076", + "pull_request:44074|pull_request:44101", + "pull_request:44056|pull_request:44161", + "pull_request:43997|pull_request:44018", + "pull_request:43997|pull_request:44161", + "pull_request:44074|pull_request:44161", + "pull_request:44018|pull_request:44161", + "pull_request:43997|pull_request:44071", + "pull_request:44010|pull_request:44071", + "pull_request:44010|pull_request:44073", + "pull_request:44018|pull_request:44129", + "pull_request:43997|pull_request:44074", + "pull_request:43997|pull_request:44129", + "pull_request:44071|pull_request:44129", + "pull_request:44010|pull_request:44018", + "pull_request:43996|pull_request:44161", + "pull_request:43996|pull_request:44017", + "pull_request:44010|pull_request:44044", + "pull_request:44044|pull_request:44074", + "pull_request:44066|pull_request:44722", + "pull_request:44085|pull_request:44722", + "pull_request:44085|pull_request:44129", + "pull_request:43996|pull_request:44007", + "pull_request:43996|pull_request:44129", + "pull_request:44044|pull_request:44129", + "pull_request:44044|pull_request:44154" + ], + "prepared_review_unit_hash": "ed899cbd3b0969dea734fc8f4dd19142f25c553c68c0ab31386b888d04ee1bbc", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 + }, + "budget": { + "node_count": 27, + "item_count": 27, + "soft_pair_count": 72, + "serialized_chars": 75095, + "estimated_input_tokens": 18774, + "estimated_eval_tokens": 37804 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "This cluster is mostly a theme cluster of per-model output-tracing refactors, not a single duplicate bug. The only strong duplicate-like matches are within the same model family/file (RegNet, GPT-Neo, GPT-J/CodeGen); cross-model pairs should be rejected.", + "confidence": 0.9, + "canonical_issue_reason": null, + "canonical_pr_reason": null, + "best_issue_reason": null, + "best_pr_reason": "PR 44722 is the strongest representative of the GPT-J/CodeGen subset and the most complete late-stage refactor among the near-duplicates, but it does not represent the whole heterogeneous cluster.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43996", + "right": "pull_request:44085", + "accept": false, + "reason": "Different model families; same umbrella output-tracing theme is not enough." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44044", + "accept": false, + "reason": "CVT/FNet vs DeBERTa are unrelated model changes." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44085", + "accept": true, + "reason": "Both are GPT-J output-tracing refactors in the same code path; 44066 is essentially a broader superset." + }, + { + "left": "pull_request:44007", + "right": "pull_request:44072", + "accept": false, + "reason": "RegNet/ResNet family vs EfficientNet; different concrete changes." + }, + { + "left": "pull_request:44072", + "right": "pull_request:44722", + "accept": false, + "reason": "EfficientNet and GPT-J/CodeGen are unrelated." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44072", + "accept": false, + "reason": "GPT-J/CodeGen vs EfficientNet are different model refactors." + }, + { + "left": "pull_request:44013", + "right": "pull_request:44044", + "accept": false, + "reason": "MobileNetV2 and DeBERTa are unrelated." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-J/CodeGen vs MGP-STR are unrelated changes." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-J/CodeGen vs MPT are unrelated." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44068", + "accept": true, + "reason": "Same GPT-Neo output-tracing refactor in the same model/file; one PR is a superset-like variant." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-J/CodeGen and GPT-Neo are different model implementations." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44066", + "accept": false, + "reason": "GPT-Neo vs GPT-J/CodeGen are different concrete model paths." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44722", + "accept": false, + "reason": "ResNet and GPT-J/CodeGen are unrelated." + }, + { + "left": "pull_request:44086", + "right": "pull_request:44722", + "accept": false, + "reason": "MGP-STR and GPT-J/CodeGen are unrelated." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44085", + "accept": false, + "reason": "ResNet and GPT-J are unrelated." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44019", + "accept": false, + "reason": "GPT-Neo and ResNet are unrelated model refactors." + }, + { + "left": "pull_request:44019", + "right": "pull_request:44071", + "accept": false, + "reason": "ResNet and MPT are unrelated." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44116", + "accept": false, + "reason": "GPT-Neo and Flaubert/XLM are unrelated." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44722", + "accept": false, + "reason": "GPT-Neo and GPT-J/CodeGen are different model families." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44059", + "accept": false, + "reason": "GPT-Neo and GPT-2 are different concrete changes." + }, + { + "left": "pull_request:44068", + "right": "pull_request:44086", + "accept": false, + "reason": "GPT-Neo and MGP-STR are unrelated." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44068", + "accept": false, + "reason": "GPT-2 and GPT-Neo are separate model implementations." + }, + { + "left": "pull_request:44059", + "right": "pull_request:44071", + "accept": false, + "reason": "GPT-2 and MPT are unrelated." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44161", + "accept": false, + "reason": "VisualBert and LongT5 are unrelated." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44017", + "accept": false, + "reason": "RegNet and Segformer are different model code paths." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44056", + "accept": false, + "reason": "RegNet and MPNet are unrelated." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are unrelated." + }, + { + "left": "pull_request:44074", + "right": "pull_request:44101", + "accept": false, + "reason": "TextNet and Flaubert/XLM are unrelated." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44161", + "accept": false, + "reason": "MPNet and LongT5 are unrelated." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44018", + "accept": false, + "reason": "RegNet and GPT-Neo are different model refactors." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44161", + "accept": false, + "reason": "RegNet and LongT5 are unrelated." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44071", + "accept": false, + "reason": "RegNet and MPT are unrelated." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44071", + "accept": false, + "reason": "SqueezeBert and MPT are unrelated." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44073", + "accept": false, + "reason": "SqueezeBert and VisualBert are unrelated." + }, + { + "left": "pull_request:44018", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-Neo and SpeechT5 are unrelated." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44074", + "accept": false, + "reason": "RegNet and TextNet are unrelated." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44129", + "accept": false, + "reason": "RegNet and SpeechT5 are unrelated." + }, + { + "left": "pull_request:44071", + "right": "pull_request:44129", + "accept": false, + "reason": "MPT and SpeechT5 are unrelated." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44018", + "accept": false, + "reason": "SqueezeBert and GPT-Neo are unrelated." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44161", + "accept": false, + "reason": "CVT/FNet and LongT5 are unrelated." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44017", + "accept": false, + "reason": "CVT/FNet and Segformer are unrelated." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44044", + "accept": false, + "reason": "SqueezeBert and DeBERTa are unrelated." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44074", + "accept": false, + "reason": "DeBERTa and TextNet are unrelated." + }, + { + "left": "pull_request:44066", + "right": "pull_request:44722", + "accept": true, + "reason": "Same GPT-J/CodeGen output-tracing refactor; 44722 is the broader later variant." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44722", + "accept": true, + "reason": "Both touch the same GPT-J output-tracing code path; 44722 subsumes the narrower GPT-J-only variant." + }, + { + "left": "pull_request:44085", + "right": "pull_request:44129", + "accept": false, + "reason": "GPT-J and SpeechT5 are unrelated." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44007", + "accept": false, + "reason": "CVT/FNet vs ResNet/RegNet/RT-DETR are different model changes." + }, + { + "left": "pull_request:43996", + "right": "pull_request:44129", + "accept": false, + "reason": "CVT/FNet and SpeechT5 are unrelated." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44129", + "accept": false, + "reason": "DeBERTa and SpeechT5 are unrelated." + }, + { + "left": "pull_request:44044", + "right": "pull_request:44154", + "accept": false, + "reason": "DeBERTa and VITS are unrelated." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet and stays conservative: it correctly treats this as a heterogeneous output-tracing theme cluster rather than one bug, and the only accepted soft edges are the same-family/file overlaps that look mergeable (GPT-Neo and GPT-J/CodeGen supersets). Cross-model rejections are well supported by the diffs and titles." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 14, + "component_count": 35, + "review_unit_index": 2, + "review_unit_count": 2, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43996-28", + "nodes": [ + "pull_request:43997", + "pull_request:44010", + "pull_request:44024", + "pull_request:44056", + "pull_request:44073", + "pull_request:44074", + "pull_request:44076", + "pull_request:44129" + ], + "soft_pairs": [ + "pull_request:44024|pull_request:44076", + "pull_request:44073|pull_request:44074", + "pull_request:44056|pull_request:44073", + "pull_request:44056|pull_request:44074", + "pull_request:43997|pull_request:44010", + "pull_request:44010|pull_request:44129" + ], + "prepared_review_unit_hash": "a5365c581e2b86c4fe65d204e86afd9edc0c9322bc291242dd230a400aaa510a", + "component_budget": { + "node_count": 28, + "item_count": 28, + "soft_pair_count": 78, + "serialized_chars": 84900, + "estimated_input_tokens": 21225, + "estimated_eval_tokens": 42706 + }, + "budget": { + "node_count": 8, + "item_count": 8, + "soft_pair_count": 6, + "serialized_chars": 17126, + "estimated_input_tokens": 4282, + "estimated_eval_tokens": 8820 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": true, + "analyst_result": { + "summary": "All items are separate model-specific refactors to standardized output tracing/capture; they share a theme but not the same underlying bug or change, so they should not be deduplicated together.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 44129 (SpeechT5) is the strongest standalone candidate because it has an explicit issue target, a concrete layer-level output-signature change, and a fuller implementation path than the other model migrations.", + "best_issue_reason": null, + "best_pr_reason": "PR 44129 is the best representative of the cluster as a concrete, issue-linked example of the output-capture refactor pattern, though it is not a duplicate of the others.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44024", + "right": "pull_request:44076", + "accept": false, + "reason": "Both are output-tracing migrations, but they target different models (FocalNet vs ImageGPT) and different code paths; this is the same refactor pattern, not the same change." + }, + { + "left": "pull_request:44073", + "right": "pull_request:44074", + "accept": false, + "reason": "Both standardize output collection, but one is VisualBert and the other is TextNet; they are unrelated model-specific edits, not one shared bug fix." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44073", + "accept": false, + "reason": "MPNet and VisualBert have similar decorator-based output tracing refactors, but they modify different architectures and hooks, so they are not the same underlying change." + }, + { + "left": "pull_request:44056", + "right": "pull_request:44074", + "accept": false, + "reason": "These are separate model migrations with different forward/output plumbing; similarity is only at the refactor pattern level." + }, + { + "left": "pull_request:43997", + "right": "pull_request:44010", + "accept": false, + "reason": "RegNet and SqueezeBert changes both touch output recording infrastructure, but they do not fix the same concrete code path or bug." + }, + { + "left": "pull_request:44010", + "right": "pull_request:44129", + "accept": false, + "reason": "SqueezeBert and SpeechT5 are distinct model implementations with different output contracts; the overlap is only the standardized-output effort." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 15, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-43998-14", + "nodes": [ + "pull_request:43998", + "pull_request:43999", + "pull_request:44000", + "pull_request:44001", + "pull_request:44002", + "pull_request:44003", + "pull_request:44004", + "pull_request:44025", + "pull_request:44026", + "pull_request:44027", + "pull_request:44028", + "pull_request:44029", + "pull_request:44030", + "pull_request:46438" + ], + "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:44025", + "pull_request:44000|pull_request:44026", + "pull_request:44001|pull_request:44002", + "pull_request:44001|pull_request:44025", + "pull_request:44001|pull_request:44026", + "pull_request:44002|pull_request:44029", + "pull_request:44002|pull_request:44025", + "pull_request:44001|pull_request:44003", + "pull_request:44002|pull_request:44003", + "pull_request:44000|pull_request:44003", + "pull_request:44003|pull_request:44028", + "pull_request:44003|pull_request:44004", + "pull_request:43998|pull_request:44004", + "pull_request:44001|pull_request:44004", + "pull_request:44002|pull_request:44004", + "pull_request:44000|pull_request:44004", + "pull_request:44025|pull_request:44026", + "pull_request:44025|pull_request:44027", + "pull_request:44025|pull_request:44028", + "pull_request:44025|pull_request:44029", + "pull_request:44025|pull_request:44030", + "pull_request:44026|pull_request:44029", + "pull_request:44026|pull_request:44027", + "pull_request:44026|pull_request:44028", + "pull_request:44026|pull_request:44030", + "pull_request:44027|pull_request:44028", + "pull_request:44027|pull_request:44029", + "pull_request:44027|pull_request:44030", + "pull_request:44028|pull_request:44029", + "pull_request:44028|pull_request:44030", + "pull_request:44029|pull_request:44030", + "pull_request:44027|pull_request:46438", + "pull_request:44000|pull_request:46438", + "pull_request:44029|pull_request:46438", + "pull_request:44028|pull_request:46438", + "pull_request:43998|pull_request:46438" + ], + "prepared_review_unit_hash": "c0ee4af05efb8b2e8084b0884b7fc1eb6656ed4752fbd2e4bfae6f6b3fc3d1a8", + "component_budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 48, + "serialized_chars": 45107, + "estimated_input_tokens": 11277, + "estimated_eval_tokens": 22810 + }, + "budget": { + "node_count": 14, + "item_count": 14, + "soft_pair_count": 48, + "serialized_chars": 45107, + "estimated_input_tokens": 11277, + "estimated_eval_tokens": 22810 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "This cluster is mostly a batch of similar-but-separate output-tracing refactors across different model implementations, plus one unrelated modular-conversion speedup PR. They share a pattern, but they are not the same concrete fix/change and should not be merged as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "No issue is present; as a representative PR, 44003 is the broadest of the refactor series because it updates two related Mamba-family model files and introduces the tracing/capture plumbing used by the rest of the set.", + "best_issue_reason": null, + "best_pr_reason": "PR 44003 is the best representative of the cluster\u2019s main theme: an output-tracing refactor in closely related model code. It is more central than the single-model PRs, while 46438 is unrelated.", + "soft_edge_verdicts": [ + { + "left": "pull_request:43998", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model families/files (timm_backbone vs upernet); same refactor pattern, not the same concrete change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44000", + "accept": false, + "reason": "Different models and code paths; both refactor output handling, but not a single duplicate fix." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44001", + "accept": false, + "reason": "Different models/files; similar tracing cleanup, but not the same bug or mergeable PR." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44003", + "accept": false, + "reason": "Different subsystems/models; one is timm_backbone, the other Mamba-family tracing changes." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44029", + "accept": false, + "reason": "Different architectures and files; only the high-level refactor theme matches." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44001", + "accept": false, + "reason": "Different models (MobileNetV1 vs UnivNet); shared output-tracing motif is too broad." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model code paths; not the same underlying fix." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44025", + "accept": false, + "reason": "Different models/files; both add can_return_tuple-style tracing, but they are separate changes." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44026", + "accept": false, + "reason": "Different encoder-decoder model wrappers; not one concrete duplicate change." + }, + { + "left": "pull_request:43999", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model families; same broad refactor category only." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44001", + "accept": false, + "reason": "Vision-text dual encoder vs UnivNet; too different to be the same fix." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44002", + "accept": false, + "reason": "Different model implementations; not mergeable as one PR." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44025", + "accept": false, + "reason": "Different tasks/models; both are output-tracing refactors, but not duplicates." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model wrappers and files; same refactor pattern only." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44002", + "accept": false, + "reason": "Unrelated model code paths; not the same concrete bug/change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44025", + "accept": false, + "reason": "Different model families and files; too broad to treat as duplicate." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44026", + "accept": false, + "reason": "Different model wrappers; same output-tracing theme, but separate PRs." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model architectures and files; not the same underlying code-path fix." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44025", + "accept": false, + "reason": "Different models; both refactor output handling, but not one shared change." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families and code paths; not a duplicate PR pair." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44003", + "accept": false, + "reason": "Different models/subsystems; only superficial similarity." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44003", + "accept": false, + "reason": "Different model families; same general tracing motif is insufficient." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model code paths; both are refactors, not the same fix." + }, + { + "left": "pull_request:44003", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model domains (Mamba vs CodeGen); not mergeable as one PR." + }, + { + "left": "pull_request:43998", + "right": "pull_request:44004", + "accept": false, + "reason": "Different models/files; only the output-tracing refactor pattern overlaps." + }, + { + "left": "pull_request:44001", + "right": "pull_request:44004", + "accept": false, + "reason": "Different model implementations; not the same concrete bug or change." + }, + { + "left": "pull_request:44002", + "right": "pull_request:44004", + "accept": false, + "reason": "Different models and code paths; too broad to dedupe." + }, + { + "left": "pull_request:44000", + "right": "pull_request:44004", + "accept": false, + "reason": "Different architectures; same refactor theme but not the same PR." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44026", + "accept": false, + "reason": "Related refactor style, but different model wrapper files and behavior." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model wrappers; not the same underlying change." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44028", + "accept": false, + "reason": "Different tasks/models; output tracing cleanup alone is too broad." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model code paths; same pattern, not duplicate." + }, + { + "left": "pull_request:44025", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families and files; not mergeable as one PR." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44029", + "accept": false, + "reason": "Different seq2seq wrapper files; only the refactor motif matches." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44027", + "accept": false, + "reason": "Different model wrappers and code paths; not the same change." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model domains; not a duplicate fix." + }, + { + "left": "pull_request:44026", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model code paths; shared tracing refactor is too generic." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44028", + "accept": false, + "reason": "Different model families (speech encoder-decoder vs SuperPoint); not the same concrete bug/change." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model implementations; not mergeable as one PR." + }, + { + "left": "pull_request:44027", + "right": "pull_request:44030", + "accept": false, + "reason": "Different architectures; same high-level refactor theme only." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44029", + "accept": false, + "reason": "Different model code paths; not a duplicate." + }, + { + "left": "pull_request:44028", + "right": "pull_request:44030", + "accept": false, + "reason": "Different model families; output-tracing cleanup is not enough to combine them." + }, + { + "left": "pull_request:44029", + "right": "pull_request:44030", + "accept": false, + "reason": "Different models/subsystems; similar refactor category, but separate changes." + }, + { + "left": "pull_request:44027", + "right": "pull_request:46438", + "accept": false, + "reason": "Unrelated changes: speech encoder-decoder output tracing vs modular-conversion process-pool reuse." + }, + { + "left": "pull_request:44000", + "right": "pull_request:46438", + "accept": false, + "reason": "Unrelated model refactor vs build/tooling performance change." + }, + { + "left": "pull_request:44029", + "right": "pull_request:46438", + "accept": false, + "reason": "Completely different code paths; one is RWKV tracing, the other a modular conversion utility optimization." + }, + { + "left": "pull_request:44028", + "right": "pull_request:46438", + "accept": false, + "reason": "Different domains and purposes; not the same underlying change." + }, + { + "left": "pull_request:43998", + "right": "pull_request:46438", + "accept": false, + "reason": "Model refactor and tooling optimization are unrelated." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 16, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44662-2", + "nodes": [ + "pull_request:44662", + "pull_request:45401" + ], + "soft_pairs": [ + "pull_request:44662|pull_request:45401" + ], + "prepared_review_unit_hash": "84c3470a75dea818b4949d854d0fb39aa539da9bcbc3c95b489e1c3b842fd657", + "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: one implements PenguinVL, the other adds Voxtral TTS. The overlap is limited to standard auto-registration/docs scaffolding, not the same underlying change.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "No single canonical PR: the changes target different models, different code paths, and would not be merged as one patch.", + "best_issue_reason": null, + "best_pr_reason": "PR 45401 is the cleaner representative if one must be picked because it has an explicit issue target and a narrower, self-contained Voxtral TTS scope.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44662", + "right": "pull_request:45401", + "accept": false, + "reason": "Different model integrations (PenguinVL vs Voxtral TTS); shared files are generic registration/docs plumbing, not the same bug or feature." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 17, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-44979-4", + "nodes": [ + "pull_request:44979", + "pull_request:45363", + "pull_request:45666", + "pull_request:46339" + ], + "soft_pairs": [ + "pull_request:44979|pull_request:45363", + "pull_request:45363|pull_request:46339", + "pull_request:45666|pull_request:46339" + ], + "prepared_review_unit_hash": "bc9a53cba695ef23ec5e78cb64843558cd22457fb2effde61f9db1ff1f25677e", + "component_budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 3, + "serialized_chars": 9633, + "estimated_input_tokens": 2409, + "estimated_eval_tokens": 5074 + }, + "budget": { + "node_count": 4, + "item_count": 4, + "soft_pair_count": 3, + "serialized_chars": 9633, + "estimated_input_tokens": 2409, + "estimated_eval_tokens": 5074 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All four items are closely related kernel-fusion PRs, but they are not all the same change. The clearest canonical PR is the latest, most complete n-to-1 fusion refinement (46339).", + "confidence": 0.79, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46339 is the most recent and most comprehensive refinement of the n-to-1 KernelConfig fusion work, with broader edits in the same loading/integration path and tests, so it is the best representative of the cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR 46339 best captures the cluster\u2019s main change set: it extends and simplifies the same KernelConfig-based n-to-1 fusion mechanism across core loading, hub kernels, modeling utils, and tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:44979", + "right": "pull_request:45363", + "accept": false, + "reason": "Related feature area, but 44979 introduces a new Module Fusion API while 45363 implements n-to-1 KernelConfig fusion; these are adjacent foundations, not the same concrete PR-level change." + }, + { + "left": "pull_request:45363", + "right": "pull_request:46339", + "accept": true, + "reason": "Both target the same concrete n-to-1 KernelConfig fusion path in the model-loading/integration stack, and 46339 reads like a direct refinement of the same feature work." + }, + { + "left": "pull_request:45666", + "right": "pull_request:46339", + "accept": true, + "reason": "These appear to be successive revisions of the same n-to-1 KernelConfig fusion implementation, touching the same code paths and sharing the same underlying feature goal." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: 46339 is the latest and broadest n-to-1 KernelConfig fusion refinement, and the soft-edge verdicts stay conservative. The 44979 vs 45363 rejection is appropriately cautious, while the 45363\u219446339 and 45666\u219446339 accepts are supported by overlapping filenames and the shared feature path." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 18, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45086-2", + "nodes": [ + "pull_request:45086", + "pull_request:45317" + ], + "soft_pairs": [ + "pull_request:45086|pull_request:45317" + ], + "prepared_review_unit_hash": "335c8e09379e95a12adb80a8a99579ade430794451ec62563bb49c1380f6064b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4414, + "estimated_input_tokens": 1104, + "estimated_eval_tokens": 2464 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs fix the same AttributeError in `_patch_mistral_regex` by changing the pre-tokenizer access from `backend_tokenizer.pre_tokenizer` to `tokenizer.pre_tokenizer`; 45317 is the more complete version because it adds a test and appears to be the later refinement of the same fix.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45317 is the better canonical PR because it targets the same bug as 45086 but includes a test change and looks like the more complete, later fix.", + "best_issue_reason": null, + "best_pr_reason": "PR 45317 is the best representative because it directly fixes the AttributeError and adds coverage, making it the clearest and most complete implementation of the change.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45086", + "right": "pull_request:45317", + "accept": true, + "reason": "Same concrete bug fix in the same code path and same issue target; 45317 is essentially the follow-up/expanded version of 45086." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 19, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45115-2", + "nodes": [ + "pull_request:45115", + "pull_request:45569" + ], + "soft_pairs": [ + "pull_request:45115|pull_request:45569" + ], + "prepared_review_unit_hash": "59007ab6f309b9a856c6f05ff28270fdad6ddc0c19e91dc5512750923e3f1247", + "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": "These PRs are related NemotronH work, but they target different end states: one refactors the existing NemotronH implementation to inherit GraniteMoeHybrid-style conversion/mapping, while the other splits NemotronH into dedicated dense/sparse variants and turns nemotron_h into a compatibility dispatcher. They overlap in shared model files, but not in the same concrete fix.", + "confidence": 0.82, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45569 is the better cluster representative because it is the broader, newer architectural change that introduces the dense/sparse split and compatibility layer, whereas #45115 is a narrower refactor/conversion mapping update.", + "best_issue_reason": null, + "best_pr_reason": "PR #45569 is the strongest standalone PR to keep: it contains the more complete user-facing model restructuring and associated docs/tests, making it the most meaningful canonical artifact for NemotronH changes.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45115", + "right": "pull_request:45569", + "accept": false, + "reason": "They overlap on NemotronH internals, but #45115 is a refactor/conversion-mapping change and #45569 is a model-splitting/dispatcher change. They are related work, not the same concrete bug or 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": 20, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45332-2", + "nodes": [ + "pull_request:45332", + "pull_request:45333" + ], + "soft_pairs": [ + "pull_request:45332|pull_request:45333" + ], + "prepared_review_unit_hash": "f5e7416d7752738b0b1ea74cb13cb2dca999f49a3d95fbfc2645f0c111be5e23", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5466, + "estimated_input_tokens": 1367, + "estimated_eval_tokens": 2990 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5466, + "estimated_input_tokens": 1367, + "estimated_eval_tokens": 2990 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are related feature PRs for heterogeneous model support, but they are not duplicates: one adds per-layer configuration, while the other adds downstream modeling/runtime support that consumes it.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45332 is the broader end-to-end change: it covers heterogeneous modeling/runtime behavior across cache, masking, modeling, and integrations, so it is the better representative of the cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR 45332 best captures the underlying feature work because it includes the concrete model-path changes that make heterogeneous configs actionable; 45333 is a narrower foundational config-only piece.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45332", + "right": "pull_request:45333", + "accept": false, + "reason": "Related but not the same concrete change: 45333 adds per-layer config support, while 45332 adds model/runtime support that uses heterogeneous config. They could be part of the same feature area, but they are not a duplicate fix and would not naturally be merged as one PR without broadening scope." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 21, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45415-2", + "nodes": [ + "pull_request:45415", + "pull_request:45425" + ], + "soft_pairs": [ + "pull_request:45415|pull_request:45425" + ], + "prepared_review_unit_hash": "13b486fabb43328c791425d13f652cd95991b6a8f310b881863ccc113a414300", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5513, + "estimated_input_tokens": 1379, + "estimated_eval_tokens": 3014 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs are typing-related, but they are not the same change: #45415 is a broad type-checking sweep across many `src/transformers/*` files plus `pyproject.toml`, while #45425 adds a few concrete typing protocols/aliases in `_typing.py` and updates a small set of consumers. The overlap is thematic, not a duplicate code-path fix.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "#45425 is the more concrete, self-contained typing change; #45415 is a much broader repository-wide type-checking cleanup rather than the same underlying fix.", + "best_issue_reason": null, + "best_pr_reason": "#45425 best represents the cluster because it introduces specific typing abstractions and targeted call-site updates, whereas #45415 is a broad static-analysis sweep with many unrelated file edits.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45415", + "right": "pull_request:45425", + "accept": false, + "reason": "Shared typing infrastructure and a few filenames do not make these the same PR: one is a wide pyright/type-ignore sweep, the other is a focused protocol/alias addition for model utilities and PEFT integration." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 22, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45513-2", + "nodes": [ + "pull_request:45513", + "pull_request:46084" + ], + "soft_pairs": [ + "pull_request:45513|pull_request:46084" + ], + "prepared_review_unit_hash": "b37d389367a3ec7581d2b4a2ecd975d54696e7949b19b279b055793904beb97f", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4858, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4858, + "estimated_input_tokens": 1215, + "estimated_eval_tokens": 2686 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated open pull requests fixing different cache-related forward-pass bugs in different model families: Qwen3.5/Qwen3.5-MoE/Qwen3-Next linear attention vs Mamba2 CUDA-kernel cache handling.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46084 is the clearest representative: it has an explicit issue target, a focused fix for a specific seq_len>1 cache path in Mamba2, and a tight file set.", + "best_issue_reason": null, + "best_pr_reason": "PR #46084 is the best standalone representative because it is narrowly scoped, explicitly tied to issue #46032, and fixes one concrete code path. PR #45513 is broader and spans several model variants, but it addresses a different bug.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45513", + "right": "pull_request:46084", + "accept": false, + "reason": "They touch different models and different cache bugs. #45513 fixes multi-token cached forward in Qwen3.5/Qwen3-Next/Olmo Hybrid linear attention, while #46084 fixes Mamba2 CUDA-kernel use_cache behavior when seq_len > 1. They are not the same underlying 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": 23, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45548-2", + "nodes": [ + "pull_request:45548", + "pull_request:45662" + ], + "soft_pairs": [ + "pull_request:45548|pull_request:45662" + ], + "prepared_review_unit_hash": "84dc521abaf91e273548ca86d8610fdb89da22c647216d9e5d52e0afcbf2d946", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4508, + "estimated_input_tokens": 1127, + "estimated_eval_tokens": 2510 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two PRs are thematically related (expert parallelism loading), but they fix different bugs: one preserves/recognizes distributed_config for EP+DeepSpeed ZeRO-3 loading, while the other prevents expert parameters from being overwritten during EP+FSDP2 sharded loading. They should not be treated as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45662 is the more concrete representative of the cluster because it fixes a specific loading overwrite path in sharded/distributed parameter handling, with a clear code-path bug and targeted remediation.", + "best_issue_reason": null, + "best_pr_reason": "PR 45662 is the strongest standalone PR because it addresses a clearly defined failure mode (expert parameters being silently overwritten during post-shard loading) and is less of a supporting/configuration fix than 45548.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45548", + "right": "pull_request:45662", + "accept": false, + "reason": "Both concern EP loading, but 45548 is about serializing/discovering distributed_config for accelerate/ZeRO-3, while 45662 fixes a different overwrite bug in post-shard parameter assignment under FSDP2. Different concrete code paths, so not a duplicate." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 24, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45639-2", + "nodes": [ + "pull_request:45639", + "pull_request:45645" + ], + "soft_pairs": [ + "pull_request:45639|pull_request:45645" + ], + "prepared_review_unit_hash": "26423451dc28a322605bea862c8af295d042699d3c1bbc44dc3726bb26d812ca", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4589, + "estimated_input_tokens": 1148, + "estimated_eval_tokens": 2552 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address xdist-safe handling for captured_info/debug artifact files in the same testing utility path; 45645 appears to subsume the narrower fix in 45639 with additional CI-side log aggregation changes.", + "confidence": 0.89, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45645 is the broader, more complete change: it includes the xdist-safe testing_utils fix from 45639 and also updates CI/workflow handling so all captured_info shards are preserved and displayed.", + "best_issue_reason": null, + "best_pr_reason": "PR 45645 is the best representative because it covers the core bug fix plus the surrounding CI/logging adjustments needed for the feature to work end-to-end.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45639", + "right": "pull_request:45645", + "accept": true, + "reason": "Same underlying xdist collision problem around captured_info output in testing_utils; 45645 largely subsumes 45639\u2019s fix and could plausibly be merged as one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 25, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45670-3", + "nodes": [ + "pull_request:45670", + "pull_request:46320", + "pull_request:46401" + ], + "soft_pairs": [ + "pull_request:45670|pull_request:46320", + "pull_request:45670|pull_request:46401", + "pull_request:46320|pull_request:46401" + ], + "prepared_review_unit_hash": "5f9b9753300010c2bf6af320d15660043613069effe48bdf600e609a2e014d56", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7839, + "estimated_input_tokens": 1960, + "estimated_eval_tokens": 4176 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 7839, + "estimated_input_tokens": 1960, + "estimated_eval_tokens": 4176 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are three distinct pull requests: one adds an AutoModel mapping for GLM ASR, and the other two add attn_implementation dispatch support for different model families. Despite similar wording, the attn_implementation PRs touch different code paths and model sets, so they should not be merged as duplicates.", + "confidence": 0.93, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46401 is the broadest and most substantial change in the set, covering multiple model families and tests/docs for attn_implementation dispatch support.", + "best_issue_reason": null, + "best_pr_reason": "PR 46401 is the strongest representative of the cluster because it has the largest scope and the most evidence of a concrete feature implementation, while the other PRs are either narrower or unrelated.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45670", + "right": "pull_request:46320", + "accept": false, + "reason": "Different changes: GLM ASR auto-mapping vs RecurrentGemma attn_implementation dispatch support. They do not fix the same concrete code-path problem." + }, + { + "left": "pull_request:45670", + "right": "pull_request:46401", + "accept": false, + "reason": "Different changes: GLM ASR auto-mapping vs BLIP-2/InstructBLIP attn_implementation dispatch support. No shared bug or mergeable duplicate change." + }, + { + "left": "pull_request:46320", + "right": "pull_request:46401", + "accept": false, + "reason": "Both are attn_implementation-related, but they target different model families and code paths; this is a reusable pattern, not 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": "pull_request", + "component_index": 26, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45720-2", + "nodes": [ + "pull_request:45720", + "pull_request:46259" + ], + "soft_pairs": [ + "pull_request:45720|pull_request:46259" + ], + "prepared_review_unit_hash": "e77c01ab2cba0baff6456163cc9900b254240ab3c5e6150d37957fe33a80dcb7", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7187, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 7187, + "estimated_input_tokens": 1797, + "estimated_eval_tokens": 3850 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These are two unrelated linter/modeling-rule PRs that happen to overlap on some Transformer model files. PR 45720 implements part of TRF003 for tuple-return behavior, while PR 46259 adds a different TRF018 rule about dead do_* flags in image/video processors; they should not be merged as duplicates.", + "confidence": 0.95, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46259 is the cleaner representative because it introduces one clearly scoped rule with a documented spec and targeted code updates, whereas PR 45720 is a broader partial implementation across many model classes.", + "best_issue_reason": null, + "best_pr_reason": "PR 46259 is the strongest standalone change in this cluster: it has a single, explicit modeling-rule goal and a cohesive implementation footprint.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45720", + "right": "pull_request:46259", + "accept": false, + "reason": "Different underlying changes: TRF003 tuple-return handling vs TRF018 dead do_* flag validation. The shared files reflect the same subsystem being linted, 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": "pull_request", + "component_index": 27, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45737-2", + "nodes": [ + "pull_request:45737", + "pull_request:45738" + ], + "soft_pairs": [ + "pull_request:45737|pull_request:45738" + ], + "prepared_review_unit_hash": "cc2ade73a49ec6a8883a7a24c433988645588bcd049e8aad2bce56b1b24daa1c", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4131, + "estimated_input_tokens": 1033, + "estimated_eval_tokens": 2322 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both PRs address the same MusicgenMelody cache initialization bug in `forward()`: replacing `EncoderDecoderCache(DynamicCache, DynamicCache)` with a single `DynamicCache`. PR #45737 is the more complete duplicate because it bundles the fix with a regression test, while #45738 is the fix-only variant.", + "confidence": 0.97, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #45737 is the best canonical PR because it contains the same code fix as #45738 plus a regression test, making it the most complete representation of the bug fix.", + "best_issue_reason": null, + "best_pr_reason": "PR #45737 is the strongest overall candidate: it fixes the concrete cache bug and adds a targeted regression test, whereas #45738 only changes the implementation.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45737", + "right": "pull_request:45738", + "accept": true, + "reason": "Same underlying code-path bug in `musicgen_melody` cache setup; the diffs match on the core fix and the test-only addition in #45737 could plausibly live in the same PR as #45738." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The assessment is grounded in the packet: both PRs make the same `musicgen_melody` cache initialization change, and #45737 only adds a regression test on top of that fix. The soft-edge accept is conservative enough because the diffs are mergeable into one PR for the same bugfix." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 28, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-45932-2", + "nodes": [ + "pull_request:45932", + "pull_request:46100" + ], + "soft_pairs": [ + "pull_request:45932|pull_request:46100" + ], + "prepared_review_unit_hash": "22b9dbf619b5ef92d7d2500ba4769e0aafbabe49cd357078832f9fc8b296e75a", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4300, + "estimated_input_tokens": 1075, + "estimated_eval_tokens": 2406 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "These two PRs are not duplicates: one is a broad CI/modeling change for Grafana/testing infrastructure, while the other is a narrow test-file migration PR with intentional failing assertions. They do not appear to fix the same code-path problem and would not reasonably be merged into one PR.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 45932 is the more substantial and representative change in the cluster, spanning CI wiring and multiple model files; PR 46100 is a tiny, unrelated test-only migration stub.", + "best_issue_reason": null, + "best_pr_reason": "PR 45932 is the best representative artifact because it contains the broader actual code/config changes, whereas PR 46100 is just a localized test-file edit.", + "soft_edge_verdicts": [ + { + "left": "pull_request:45932", + "right": "pull_request:46100", + "accept": false, + "reason": "Different underlying changes: 45932 updates CI secrets and multiple model implementations, while 46100 only edits a ViT test file with dummy assertions. They do not share a concrete bug or mergeable 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": 29, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46073-3", + "nodes": [ + "pull_request:46073", + "pull_request:46154", + "pull_request:46196" + ], + "soft_pairs": [ + "pull_request:46073|pull_request:46154", + "pull_request:46073|pull_request:46196", + "pull_request:46154|pull_request:46196" + ], + "prepared_review_unit_hash": "ef36305cc5c858681844e79f095bd8f00d75b3899b538200297b529640b06471", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8082, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 3, + "serialized_chars": 8082, + "estimated_input_tokens": 2021, + "estimated_eval_tokens": 4298 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All three PRs are related by the same feature theme (adding SDPA dispatch support), but they target different model families and distinct code paths, so they should not be deduplicated.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "No PR is a true duplicate of the others. If a representative must be chosen, PR #46196 is the broadest multi-model SDPA dispatch change, but it still covers separate architectures and should remain distinct.", + "best_issue_reason": null, + "best_pr_reason": "PR #46196 is the most comprehensive candidate because it spans multiple model implementations and tests, but it is not the same fix as the CTRL or T5 PRs.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46073", + "right": "pull_request:46154", + "accept": false, + "reason": "Both add SDPA dispatch support, but for different model implementations (CTRL vs T5) and different code paths; they are related feature work, not the same concrete change." + }, + { + "left": "pull_request:46073", + "right": "pull_request:46196", + "accept": false, + "reason": "CTRL SDPA dispatch is a separate model-specific implementation from the wav2vec2/SeamlessM4T changes; too broad to treat as one mergeable fix." + }, + { + "left": "pull_request:46154", + "right": "pull_request:46196", + "accept": false, + "reason": "T5 SDPA dispatch and the wav2vec2/SeamlessM4T dispatch work affect different architectures and files, so they are not duplicates of the same underlying bug or patch." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 30, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46126-2", + "nodes": [ + "pull_request:46126", + "pull_request:46269" + ], + "soft_pairs": [ + "pull_request:46126|pull_request:46269" + ], + "prepared_review_unit_hash": "6d705601f4b1c10961961768bea7491020d8ff5a072b29601a2ba7d77e749805", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5572, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5572, + "estimated_input_tokens": 1393, + "estimated_eval_tokens": 3042 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Both items are pull requests in the distributed stack, but they are not the same concrete change. PR #46126 is a focused EP + Trainer integration feature, while PR #46269 is a broad distributed branch base touching many unrelated areas. The filename overlap is not enough to treat them as duplicates.", + "confidence": 0.92, + "canonical_issue_reason": null, + "canonical_pr_reason": "No true duplicate PR pair here; if one needs to represent the cluster, #46126 is the more focused and self-contained change, whereas #46269 is a broad branch-base PR rather than the same fix.", + "best_issue_reason": null, + "best_pr_reason": "#46126 is the better representative because it has a clear, specific feature goal and a tighter code-path scope. #46269 is too broad and mixed to serve as a canonical duplicate target.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46126", + "right": "pull_request:46269", + "accept": false, + "reason": "Reject: overlapping distributed filenames, but different scope and intent. #46126 adds EP/Trainer integration; #46269 is a broad distributed branch base. They do not appear to fix the same concrete bug or implement 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": 31, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46138-6", + "nodes": [ + "pull_request:46138", + "pull_request:46432", + "pull_request:46501", + "pull_request:46502", + "pull_request:46503", + "pull_request:46504" + ], + "soft_pairs": [ + "pull_request:46138|pull_request:46504", + "pull_request:46138|pull_request:46432", + "pull_request:46138|pull_request:46503", + "pull_request:46138|pull_request:46501", + "pull_request:46138|pull_request:46502", + "pull_request:46432|pull_request:46503", + "pull_request:46432|pull_request:46501", + "pull_request:46432|pull_request:46502", + "pull_request:46432|pull_request:46504", + "pull_request:46501|pull_request:46503", + "pull_request:46501|pull_request:46502", + "pull_request:46501|pull_request:46504", + "pull_request:46502|pull_request:46503", + "pull_request:46502|pull_request:46504", + "pull_request:46503|pull_request:46504" + ], + "prepared_review_unit_hash": "e478cef77dea0d7346b24f131693ee5af22e5a3db8b3f454acd2b0c2a0dc18ee", + "component_budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14229, + "estimated_input_tokens": 3558, + "estimated_eval_tokens": 7372 + }, + "budget": { + "node_count": 6, + "item_count": 6, + "soft_pair_count": 15, + "serialized_chars": 14229, + "estimated_input_tokens": 3558, + "estimated_eval_tokens": 7372 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "All six pull requests are small, unrelated workflow/CI maintenance changes. They share only superficial similarity in title and file type, not the same underlying bug or change, so none should be clustered as duplicates.", + "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:46138", + "right": "pull_request:46432", + "accept": false, + "reason": "Different workflows and different intents: one changes self-comment CI stdin handling, the other hardens SSH runner download/secret handling." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46501", + "accept": false, + "reason": "Unrelated workflow targets; anti-slop permissions change is not the same code-path as self-comment CI command piping." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46502", + "accept": false, + "reason": "Different automation job and behavior; assign-reviewers checkout ref fix is unrelated to self-comment CI." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46503", + "accept": false, + "reason": "Benchmark workflow trigger expansion is a separate change from self-comment CI input handling." + }, + { + "left": "pull_request:46138", + "right": "pull_request:46504", + "accept": false, + "reason": "Build-image trigger parsing fix is a different workflow and problem than self-comment CI stdin passing." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46503", + "accept": false, + "reason": "Both touch CI workflows, but one is SSH/security hardening and the other is benchmark trigger logic; no shared bug or mergeable combined fix." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46501", + "accept": false, + "reason": "Permissions change in anti-slop is unrelated to SSH runner secret cleanup and checksum pinning." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46502", + "accept": false, + "reason": "Assign-reviewers checkout ref fix is a distinct workflow issue, not the SSH runner hardening change." + }, + { + "left": "pull_request:46432", + "right": "pull_request:46504", + "accept": false, + "reason": "Build-ci-docker-images command quoting fix is unrelated to SSH runner download/secret handling." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46503", + "accept": false, + "reason": "Different workflows and different behavior changes; permissions scoping vs benchmark dispatch support." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46502", + "accept": false, + "reason": "Anti-slop permission placement and assign-reviewers checkout ref are unrelated fixes." + }, + { + "left": "pull_request:46501", + "right": "pull_request:46504", + "accept": false, + "reason": "No common underlying bug: anti-slop permissions vs build-ci-docker-images shell quoting." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46503", + "accept": false, + "reason": "Assign-reviewers checkout ref fix and benchmark workflow dispatch enablement address separate jobs and problems." + }, + { + "left": "pull_request:46502", + "right": "pull_request:46504", + "accept": false, + "reason": "Distinct workflows with different failure modes; no same concrete code-path problem." + }, + { + "left": "pull_request:46503", + "right": "pull_request:46504", + "accept": false, + "reason": "Benchmark dispatch support and build-ci-docker-images quoting fix are independent workflow changes, not one duplicate 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": 32, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46152-2", + "nodes": [ + "pull_request:46152", + "pull_request:46207" + ], + "soft_pairs": [ + "pull_request:46152|pull_request:46207" + ], + "prepared_review_unit_hash": "d7d733bf9bc2c6037704fa3e05b8b0c1367ab93f4dca3fa22b013af54b18f547", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4276, + "estimated_input_tokens": 1069, + "estimated_eval_tokens": 2394 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4276, + "estimated_input_tokens": 1069, + "estimated_eval_tokens": 2394 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two open PRs both replace file-opening source introspection, but they target different modules and different behaviors: one fixes package-resource resolution in import_utils, the other changes attention-implementation detection in modeling_utils. They are related in style, not duplicates.", + "confidence": 0.94, + "canonical_issue_reason": null, + "canonical_pr_reason": "No clear canonical PR: the two PRs address separate code paths and bugs rather than one shared change.", + "best_issue_reason": null, + "best_pr_reason": "PR 46207 is the more concrete and self-contained fix, with a specific behavior change in modeling_utils and accompanying tests.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46152", + "right": "pull_request:46207", + "accept": false, + "reason": "Both avoid open()-based source access, but they fix different components and different failure modes; they would not plausibly be merged into one PR." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 33, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46173-3", + "nodes": [ + "pull_request:46173", + "pull_request:46436", + "pull_request:46474" + ], + "soft_pairs": [ + "pull_request:46436|pull_request:46474", + "pull_request:46173|pull_request:46474" + ], + "prepared_review_unit_hash": "9060d9b3132af9e84e0b2842c60ceb5c437a219f67fd887ac2ff160e19813c64", + "component_budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 + }, + "budget": { + "node_count": 3, + "item_count": 3, + "soft_pair_count": 2, + "serialized_chars": 5921, + "estimated_input_tokens": 1481, + "estimated_eval_tokens": 3218 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs are effectively duplicates of the same StaticLayer type-hint fix for #45987, while the static-cache max-len PR is unrelated and should stay separate.", + "confidence": 0.98, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46173 is the earliest instance of the exact same cache_utils type-annotation change for #45987, so it is the best anchor for the duplicate cluster.", + "best_issue_reason": null, + "best_pr_reason": "PR 46173 is the strongest representative because it is the original, exact fix for the shared issue target and matches the later duplicate nearly verbatim.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46436", + "right": "pull_request:46474", + "accept": false, + "reason": "These fix different problems: 46436 changes generation cache sizing from config, while 46474 only widens StaticLayer.get_seq_length type hints. Same broad subsystem, but not the same bug or mergeable change set." + }, + { + "left": "pull_request:46173", + "right": "pull_request:46474", + "accept": true, + "reason": "They are the same underlying fix: both edit src/transformers/cache_utils.py with the same get_seq_length return-type annotation change and the same issue target #45987." + } + ] + }, + "evaluator_result": { + "accept": true, + "feedback": "The summary is grounded in the packet: PRs 46173 and 46474 have the same issue target (#45987) and the same cache_utils type-annotation diff, so treating them as duplicates is conservative and supported. The 46436 vs 46474 verdict is also appropriately conservative because the diffs address different features/problems and are not mergeable as one bugfix PR." + }, + "evaluator_used": true, + "retried": false, + "accepted_nontrivial_soft_edge": true, + "error_kind": null, + "error_message": null + }, + { + "label": "pull_request", + "component_index": 34, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "cache", + "cache_hit": true, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46293-2", + "nodes": [ + "pull_request:46293", + "pull_request:46478" + ], + "soft_pairs": [ + "pull_request:46293|pull_request:46478" + ], + "prepared_review_unit_hash": "7bd39ae48d659fdee338d62b65f52e0c5fa39368b02b874770cb97e7816d0b13", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4443, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 4443, + "estimated_input_tokens": 1111, + "estimated_eval_tokens": 2478 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "Two PRs target the same hub-kernels area and share issue #46291, but they implement different fixes: one updates kernel repo/version metadata for kernels>=0.15.0, while the other pins LayerRepository revisions to main and adds broader config/test changes. They are related, but not the same duplicate change.", + "confidence": 0.82, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR #46478 is the better canonical PR because it is broader and more complete: it updates multiple hub-kernel references, touches kernel config, and includes test coverage, making it the stronger representative of the fix set.", + "best_issue_reason": null, + "best_pr_reason": "PR #46478 is the best overall PR in the cluster because it addresses the problem more comprehensively and is more suitable as the merge target for the shared hub-kernel compatibility work.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46293", + "right": "pull_request:46478", + "accept": false, + "reason": "Both mention the same issue and file, but the concrete changes differ: #46293 adjusts repo/version compatibility metadata, while #46478 pins revisions to main and extends config/tests. They are related fixes, not the same PR-worthy code 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": 35, + "component_count": 35, + "review_unit_index": 1, + "review_unit_count": 1, + "status": "reviewed", + "reason": null, + "source": "llm", + "cache_hit": false, + "model": "gpt-5.4-mini?service_tier=flex", + "cluster_id": "cluster-46376-2", + "nodes": [ + "pull_request:46376", + "pull_request:46417" + ], + "soft_pairs": [ + "pull_request:46376|pull_request:46417" + ], + "prepared_review_unit_hash": "645314fa075543245856d8a30a2f2e83744a37df47f010f857e4bc7e8909e10b", + "component_budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5960, + "estimated_input_tokens": 1490, + "estimated_eval_tokens": 3236 + }, + "budget": { + "node_count": 2, + "item_count": 2, + "soft_pair_count": 1, + "serialized_chars": 5960, + "estimated_input_tokens": 1490, + "estimated_eval_tokens": 3236 + }, + "overflow_policy": "truncate_then_skip", + "trimmed": false, + "aggressively_trimmed": false, + "split": false, + "analyst_result": { + "summary": "The two pull requests are unrelated: one adds multimodal support to continuous batching, while the other adds a new HunYuan VL model implementation and docs. They do not look like the same underlying fix or change set.", + "confidence": 0.96, + "canonical_issue_reason": null, + "canonical_pr_reason": "PR 46417 is the better representative artifact because it is a concrete, self-contained model addition with code, docs, and tests; PR 46376 is a draft, broad infrastructure feature and not a duplicate of it.", + "best_issue_reason": null, + "best_pr_reason": "PR 46417 is the strongest standalone PR in the cluster: it has a clearly defined scope and substantial implementation coverage. PR 46376 is broader, still draft, and targets a different subsystem.", + "soft_edge_verdicts": [ + { + "left": "pull_request:46376", + "right": "pull_request:46417", + "accept": false, + "reason": "Different work items and code paths: continuous batching multimodal infrastructure versus adding the HunYuan VL model package, configs, processors, and tests. They are not plausibly mergeable into one PR as duplicates." + } + ] + }, + "evaluator_result": null, + "evaluator_used": false, + "retried": false, + "accepted_nontrivial_soft_edge": false, + "error_kind": null, + "error_message": null + } + ] +} diff --git a/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/manifest.json b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..03734a8775e147452b7f082772a8358b2a094a6e --- /dev/null +++ b/snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/manifest.json @@ -0,0 +1,14 @@ +{ + "analysis_id": "hybrid-model-20260608t180018z", + "artifacts": { + "hybrid": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "model": null, + "published_at": "2026-06-08T18:11:35Z", + "repo": "huggingface/transformers", + "schema_version": 1, + "snapshot_id": "20260608T180018Z", + "variant": "hybrid" +} diff --git a/snapshots/20260608T180018Z/manifest.json b/snapshots/20260608T180018Z/manifest.json index 5b64170a68c87d32eb7c01054b1de7541971e807..3f5a2bb65c309f30e7de382737fc6b93dcdc2062 100644 --- a/snapshots/20260608T180018Z/manifest.json +++ b/snapshots/20260608T180018Z/manifest.json @@ -32,6 +32,24 @@ "timeline_events": 1002 }, "extracted_at": "2026-06-08T18:00:18Z", + "published_analysis": { + "canonical_analysis_id": "hybrid-model-20260608t180018z", + "runs": { + "hybrid-model-20260608t180018z": { + "analysis_id": "hybrid-model-20260608t180018z", + "artifacts": { + "hybrid": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.json", + "hybrid_reviews": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/analysis-report-hybrid.llm-reviews.json" + }, + "channel": "canonical", + "manifest_path": "snapshots/20260608T180018Z/analysis-runs/hybrid-model-20260608t180018z/manifest.json", + "model": null, + "published_at": "2026-06-08T18:11:35Z", + "variant": "hybrid" + } + }, + "schema_version": 1 + }, "repo": "huggingface/transformers", "snapshot_id": "20260608T180018Z", "watermark": {